MCP Agent Mail
"It's like gmail for your coding agents!"
A mail-like coordination layer for coding agents, exposed as an HTTP-only FastMCP server. It gives agents memorable identities, an inbox/outbox, searchable message history, and voluntary file reservation "leases" to avoid stepping on each other.
Think of it as asynchronous email + directory + change-intent signaling for your agents, backed by Git (for human-auditable artifacts) and SQLite (for indexing and queries).
Status: Under active development. The design is captured in detail in project_idea_and_guide.md (start with the original prompt at the top of that file).
Why this exists
Modern projects often run multiple coding agents at once (backend, frontend, scripts, infra). Without a shared coordination fabric, agents:
- Overwrite each other's edits or panic on unexpected diffs
- Miss critical context from parallel workstreams
- Require humans to "liaison" messages across tools and teams
This project provides a lightweight, interoperable layer so agents can:
- Register a temporary-but-persistent identity (e.g., GreenCastle)
- Send/receive GitHub-Flavored Markdown messages with images
- Search, summarize, and thread conversations
- Declare advisory file reservations (leases) on files/globs to signal intent
- Inspect a directory of active agents, programs/models, and activity
It's designed for: FastMCP clients and CLI tools (Claude Code, Codex, Gemini CLI, etc.) coordinating across one or more codebases.
From Idea Spark to Shipping Swarm
If a blank repo feels daunting, follow the field-tested workflow we documented in project_idea_and_guide.md (“Appendix: From Blank Repo to Coordinated Swarm”):
- Ideate fast: Write a scrappy email-style blurb about the problem, desired UX, and any must-have stack picks (≈15 minutes).
- Promote it to a plan: Feed that blurb to GPT-5 Pro (and optionally Grok4 Heavy / Opus 4.1) until you get a granular Markdown plan, then iterate on the plan file while it’s still cheap to change. The Markdown Web Browser sample plan shows the level of detail to aim for.
- Codify the rules: Clone a tuned
AGENTS.md, add any tech-specific best-practice guides, and let Codex scaffold the repo plus Beads tasks straight from the plan. - Spin up the swarm: Launch multiple Codex panes (or any agent mix), register each identity with Agent Mail, and have them acknowledge
AGENTS.md, the plan document, and the Beads backlog before touching code. - Keep everyone fed: Reuse the canned instruction cadence from the tweet thread or, better yet, let the commercial Companion app’s Message Stacks broadcast those prompts automatically so you never hand-feed panes again.
Watch the full 23-minute walkthrough (https://youtu.be/68VVcqMEDrs?si=pCm6AiJAndtZ6u7q) to see the loop in action.
Productivity Math & Automation Loop
One disciplined hour of GPT-5 Codex—when it isn’t waiting on human prompts—often produces 10–20 “human hours” of work because the agents reason and type at machine speed. Agent Mail multiplies that advantage in two layers:
- Base OSS server: Git-backed mailboxes, advisory file reservations, Typer CLI helpers, and searchable archives keep independent agents aligned without babysitting. Every instruction, lease, and attachment is auditable.
- Companion stack (commercial): The iOS app + host automation can provision, pair, and steer heterogeneous fleets (Claude Code, Codex, Gemini CLI, etc.) from your phone using customizable Message Stacks, Human Overseer broadcasts, Beads awareness, and plan editing tools—no manual tmux choreography required. The automation closes the loop by scheduling prompts, honoring Limited Mode, and enforcing Double-Arm confirmations for destructive work.
Result: you invest 1–2 hours of human supervision, but dozens of agent-hours execute in parallel with clear audit trails and conflict-avoidance baked in.
TLDR Quickstart
One-line installer
curl -fsSL "https://raw.githubusercontent.com/Dicklesworthstone/mcp_agent_mail/main/scripts/install.sh?$(date +%s)" | bash -s -- --yes
What this does:
- Installs uv if missing and updates your PATH for this session
- Installs jq if missing (needed for safe config merging; auto-detects your package manager)
- Creates a Python 3.14 virtual environment and installs dependencies with uv
- Runs the auto-detect integration to wire up supported agent tools
- Starts the MCP HTTP server on port 8765 and prints a masked bearer token
- Creates helper scripts under
scripts/(includingrun_server_with_token.sh) - Adds an
amshell alias to your.zshrcor.bashrcfor quick server startup (just typeamin a new terminal!) - Installs/updates, verifies, and wires the Beads
bdCLI into your PATH via its official curl installer so the task planner is ready out of the box (pass--skip-beadsto opt out or install manually) - Installs/updates the Beads Viewer
bvTUI for interactive task browsing and AI-friendly robot commands (pass--skip-bvto opt out) - Prints a short on-exit summary of each setup step so you immediately know what changed
Prefer a specific location or options? Add flags like --dir <path>, --project-dir <path>, --no-start, --start-only, --port <number>, or --token <hex>.
Already have Beads or Beads Viewer installed? Append --skip-beads and/or --skip-bv to bypass automatic installation.
Starting the server in the future
After installation, you can start the MCP Agent Mail server from anywhere by simply typing:
am
That's it! The am alias (added to your .zshrc or .bashrc during installation) automatically:
- Changes to the MCP Agent Mail directory
- Runs the server startup script (which uses
uv runto handle the virtual environment) - Loads your saved bearer token from
.envand starts the HTTP server
Note: If you just ran the installer, open a new terminal or run source ~/.zshrc (or source ~/.bashrc) to load the alias.
Port conflicts? Use --port to specify a different port (default: 8765):
Install with custom port:
curl -fsSL "https://raw.githubusercontent.com/Dicklesworthstone/mcp_agent_mail/main/scripts/install.sh?$(date +%s)" | bash -s -- --port 9000 --yes
Or use the CLI command after installation: uv run python -m mcp_agent_mail.cli config set-port 9000
### If you want to do it yourself
Clone the repo, set up and install with uv in a python 3.14 venv (install uv if you don't have it already), and then run `scripts/automatically_detect_all_installed_coding_agents_and_install_mcp_agent_mail_in_all.sh`. This will automatically set things up for your various installed coding agent tools and start the MCP server on port 8765. If you want to run the MCP server again in the future, simply run `scripts/run_server_with_token.sh`:
Install uv (if you don't have it already):
```bash
curl -LsSf https://astral.sh/uv/install.sh | sh
export PATH="$HOME/.local/bin:$PATH"
# Clone the repo
git clone https://github.com/Dicklesworthstone/mcp_agent_mail
cd mcp_agent_mail
# Create a Python 3.14 virtual environment and install dependencies
# Note: If you have an older uv version, run `uv self update` first
uv python install 3.14
uv venv -p 3.14
source .venv/bin/activate
uv sync
# Detect installed coding agents, integrate, and start the MCP server on port 8765
scripts/automatically_detect_all_installed_coding_agents_and_install_mcp_agent_mail_in_all.sh
# Later, to run the MCP server again with the same token
scripts/run_server_with_token.sh
# Now, simply launch Codex-CLI or Claude Code or other agent tools in other consoles; they should have the mail tool available. See below for a ready-made chunk of text you can add to the end of your existing AGENTS.md or CLAUDE.md files to help your agents better utilize the new tools.
# Change port after installation
uv run python -m mcp_agent_mail.cli config set-port 9000
Ready-Made Blurb to Add to Your AGENTS.md or CLAUDE.md Files:
## MCP Agent Mail: coordination for multi-agent workflows
What it is
- A mail-like layer that lets coding agents coordinate asynchronously via MCP tools and resources.
- Provides identities, inbox/outbox, searchable threads, and advisory file reservations, with human-auditable artifacts in Git.
Why it's useful
- Prevents agents from stepping on each other with explicit file reservations (leases) for files/globs.
- Keeps communication out of your token budget by storing messages in a per-project archive.
- Offers quick reads (`resource://inbox/...`, `resource://thread/...`) and macros that bundle common flows.
How to use effectively
1) Same repository
- Register an identity: call `ensure_project`, then `register_agent` using this repo's absolute path as `project_key`.
- Reserve files before you edit: `file_reservation_paths(project_key, agent_name, ["src/**"], ttl_seconds=3600, exclusive=true)` to signal intent and avoid conflict.
- Communicate with threads: use `send_message(..., thread_id="FEAT-123")`; check inbox with `fetch_inbox` and acknowledge with `acknowledge_message`.
- Read fast: `resource://inbox/{Agent}?project=<abs-path>&limit=20` or `resource://thread/{id}?project=<abs-path>&include_bodies=true`.
- Tip: set `AGENT_NAME` in your environment so the pre-commit guard can block commits that conflict with others' active exclusive file reservations.
2) Across different repos in one project (e.g., Next.js frontend + FastAPI backend)
- Option A (single project bus): register both sides under the same `project_key` (shared key/path). Keep reservation patterns specific (e.g., `frontend/**` vs `backend/**`).
- Option B (separate projects): each repo has its own `project_key`; use `macro_contact_handshake` or `request_contact`/`respond_contact` to link agents, then message directly. Keep a shared `thread_id` (e.g., ticket key) across repos for clean summaries/audits.
Macros vs granular tools
- Prefer macros when you want speed or are on a smaller model: `macro_start_session`, `macro_prepare_thread`, `macro_file_reservation_cycle`, `macro_contact_handshake`.
- Use granular tools when you need control: `register_agent`, `file_reservation_paths`, `send_message`, `fetch_inbox`, `acknowledge_message`.
Common pitfalls
- "from_agent not registered": always `register_agent` in the correct `project_key` first.
- "FILE_RESERVATION_CONFLICT": adjust patterns, wait for expiry, or use a non-exclusive reservation when appropriate.
- Auth errors: if JWT+JWKS is enabled, include a bearer token with a `kid` that matches server JWKS; static bearer is used only when JWT is disabled.
Integrating with Beads (dependency-aware task planning)
Beads is a lightweight task planner (bd CLI) that complements Agent Mail by keeping status and dependencies in one place while Mail handles messaging, file reservations, and audit trails. Project: steveyegge/beads
Highlights:
- Beads owns task prioritization; Agent Mail carries the conversations and artifacts.
- Shared identifiers (e.g.,
bd-123) keep Beads issues, Mail threads, and commits aligned. - Install the
bdCLI via prebuilt release or Go build; see the repository for platform specifics.
Copy/paste blurb for agent-facing docs (leave as-is for reuse):
## Integrating with Beads (dependency-aware task planning)
Beads provides a lightweight, dependency-aware issue database and a CLI (`bd`) for selecting "ready work," setting priorities, and tracking status. It complements MCP Agent Mail's messaging, audit trail, and file-reservation signals. Project: [steveyegge/beads](https://github.com/steveyegge/beads)
Recommended conventions
- **Single source of truth**: Use **Beads** for task status/priority/dependencies; use **Agent Mail** for conversation, decisions, and attachments (audit).
- **Shared identifiers**: Use the Beads issue id (e.g., `bd-123`) as the Mail `thread_id` and prefix message subjects with `[bd-123]`.
- **Reservations**: When starting a `bd-###` task, call `file_reservation_paths(...)` for the affected paths; include the issue id in the `reason` and release on completion.
Typical flow (agents)
1) **Pick ready work** (Beads)
- `bd ready --json` → choose one item (highest priority, no blockers)
2) **Reserve edit surface** (Mail)
- `file_reservation_paths(project_key, agent_name, ["src/**"], ttl_seconds=3600, exclusive=true, reason="bd-123")`
3) **Announce start** (Mail)
- `send_message(..., thread_id="bd-123", subject="[bd-123] Start: <short title>", ack_required=true)`
4) **Work and update**
- Reply in-thread with progress and attach artifacts/images; keep the discussion in one thread per issue id
5) **Complete and release**
- `bd close bd-123 --reason "Completed"` (Beads is status authority)
- `release_file_reservations(project_key, agent_name, paths=["src/**"])`
- Final Mail reply: `[bd-123] Completed` with summary and links
Mapping cheat-sheet
- **Mail `thread_id`** ↔ `bd-###`
- **Mail subject**: `[bd-###] …`
- **File reservation `reason`**: `bd-###`
- **Commit messages (optional)**: include `bd-###` for traceability
Event mirroring (optional automation)
- On `bd update --status blocked`, send a high-importance Mail message in thread `bd-###` describing the blocker.
- On Mail "ACK overdue" for a critical decision, add a Beads label (e.g., `needs-ack`) or bump priority to surface it in `bd ready`.
Pitfalls to avoid
- Don't create or manage tasks in Mail; treat Beads as the single task queue.
- Always include `bd-###` in message `thread_id` to avoid ID drift across tools.
Prefer automation? Run uv run python -m mcp_agent_mail.cli docs insert-blurbs to scan your code directories for AGENTS.md/CLAUDE.md files and append the latest Agent Mail + Beads snippets with per-project confirmation. The installer also offers to launch this helper right after setup so you can take care of onboarding docs immediately.
Beads Viewer (bv) — AI-Friendly Task Analysis
The Beads Viewer (bv) is a fast terminal UI for Beads projects that also provides robot flags designed specifically for AI agent integration. Project: Dicklesworthstone/beads_viewer
Why bv for Agents?
While bd (Beads CLI) handles task CRUD operations, bv provides precomputed graph analytics that help agents make intelligent prioritization decisions:
- PageRank scores: Identify high-impact tasks that unblock the most downstream work
- Critical path analysis: Find the longest dependency chain to completion
- Cycle detection: Spot circular dependencies before they cause deadlocks
- Parallel track planning: Determine which tasks can run concurrently
Instead of agents parsing .beads/beads.jsonl directly or attempting to compute graph metrics (risking hallucinated results), they can call bv's deterministic robot flags and get JSON output they can trust.
Robot Flags for AI Integration
| Flag | Output | Agent Use Case |
|---|---|---|
bv --robot-help | All AI-facing commands | Discovery / capability check |
bv --robot-insights | PageRank, betweenness, HITS, critical path, cycles | Quick triage: "What's most impactful?" |
bv --robot-plan | Parallel tracks, items per track, unblocks lists | Execution planning: "What can run in parallel?" |
bv --robot-priority | Priority recommendations with reasoning + confidence | Task selection: "What should I work on next?" |
bv --robot-recipes | Available filter presets (actionable, blocked, etc.) | Workflow setup: "Show me ready work" |
bv --robot-diff --diff-since <ref> | Changes since commit/date, new/closed items, cycles | Progress tracking: "What changed?" |
Example: Agent Task Selection Workflow
# 1. Get priority recommendations with reasoning
bv --robot-priority
# Returns JSON with ranked tasks, impact scores, and confidence levels
# 2. Check what completing a task would unblock
bv --robot-plan
# Returns parallel tracks showing dependency chains
# 3. After completing work, check what changed
bv --robot-diff --diff-since "1 hour ago"
# Returns new items, closed items, cycle changes
When to Use bv vs bd
| Tool | Best For |
|---|---|
bd | Creating, updating, closing tasks; bd ready for simple "what's next" |
bv | Graph analysis, impact assessment, parallel planning, change tracking |
Rule of thumb: Use bd for task operations, use bv for task intelligence.
Integration with Agent Mail
Combine bv insights with Agent Mail coordination:
- Agent A runs
bv --robot-priority→ identifiesbd-42as highest-impact - Agent A reserves files:
file_reservation_paths(..., reason="bd-42") - Agent A announces:
send_message(..., thread_id="bd-42", subject="[bd-42] Starting high-impact refactor") - Other agents see the reservation and Mail announcement, pick different tasks
- Agent A completes, runs
bv --robot-diffto report downstream unblocks
This creates a feedback loop where graph intelligence drives coordination.
Core ideas (at a glance)
- HTTP-only FastMCP server (Streamable HTTP). No SSE, no STDIO.
- Dual persistence model:
- Human-readable markdown in a per-project Git repo for every canonical message and per-recipient inbox/outbox copy
- SQLite with FTS5 for fast search, directory queries, and file reservations/leases
- "Directory/LDAP" style queries for agents; memorable adjective+noun names
- Advisory file reservations for editing surfaces; optional pre-commit guard
- Resource layer for convenient reads (e.g.,
resource://inbox/{agent})
Typical use cases
- Multiple agents splitting a large refactor across services while staying in sync
- Frontend and backend teams of agents coordinating thread-by-thread
- Protecting critical migrations with exclusive file reservations and a pre-commit guard
- Searching and summarizing long technical discussions as threads evolve
- Discovering and linking related projects (e.g., frontend/backend) through AI-powered suggestions
Workflow FAQ
Do I still need the tmux broadcast script to “feed” every Codex pane?
No. The historical zsh loop from the tweet thread is still handy if you are running the OSS stack by itself, but the AgentMail Companion system now automates that cadence with Message Stacks. Once the companion host services are installed, you queue presets (builder loop, reviewer sweep, test focus, etc.) from the iOS app or CLI and the automation fans those instructions out to every enrolled agent—without touching tmux.
Architecture
graph LR
A[Agents]
S[Server]
G[Git repo]
Q[SQLite FTS5]
A -->|HTTP tools/resources| S
S -->|writes/reads| G
S -->|indexes/queries| Q
subgraph GitTree["Git tree"]
GI1[agents/profile.json]
GI2[agents/mailboxes/...]
GI3[messages/YYYY/\nMM/\nid.md]
GI4[file_reservations/\nsha1.json]
GA[attachments/xx/\nsha1.webp]
end
G --- GI1
G --- GI2
G --- GI3
G --- GI4
G --- GA
Web UI (human-facing mail viewer)
The server ships a lightweight, server-rendered Web UI for humans. It lets you browse projects, agents, inboxes, single messages, attachments, file reservations, and perform full-text search with FTS5 when available (with an automatic LIKE fallback).
- Where it lives: built into the HTTP server in
mcp_agent_mail.httpunder the/mailpath. - Who it's for: humans reviewing activity; agents should continue to use the MCP tools/resources API.
Launching the Web UI
Recommended (simple):
scripts/run_server_with_token.sh
# then open http://127.0.0.1:8765/mail
Advanced (manual commands):
uv run python -m mcp_agent_mail.http --host 127.0.0.1 --port 8765
# or:
uv run uvicorn mcp_agent_mail.http:build_http_app --factory --host 127.0.0.1 --port 8765
Auth notes:
- GET pages in the UI are not gated by the RBAC middleware (it classifies POSTed MCP calls only), but if you set a bearer token the separate BearerAuth middleware protects all routes by default.
- For local dev, set
HTTP_ALLOW_LOCALHOST_UNAUTHENTICATED=true(and optionallyHTTP_BEARER_TOKEN), so localhost can load the UI without headers. - Health endpoints are always open at
/health/*.
Routes and what you can do
-
/mail(Unified inbox + Projects + Related Projects Discovery)- Shows a unified, reverse-chronological inbox of recent messages across all projects with excerpts, relative timestamps, sender/recipients, and project badges.
- Below the inbox, lists all projects (slug, human name, created time) with sibling suggestions.
- Suggests likely sibling projects when two slugs appear to be parts of the same product (e.g., backend vs. frontend). Suggestions are ranked with heuristics and, when
LLM_ENABLED=true, an LLM pass across key docs (README.md,AGENTS.md, etc.). - Humans can Confirm Link or Dismiss suggestions from the dashboard. Confirmed siblings become highlighted badges but do not automatically authorize cross-project messaging; agents must still establish
AgentLinkapprovals viarequest_contact/respond_contact.
-
/mail/projects(Projects index)- Dedicated projects list view; click a project to drill in.
-
/mail/{project}(Project overview + search + agents)- Rich search form with filters:
- Scope: subject/body/both, Order: relevance or time, optional "boost subject".
- Query tokens: supports
subject:foo,body:"multi word", quoted phrases, and bare terms. - Uses FTS5 bm25 scoring when available; otherwise falls back to SQL LIKE on subject/body with your chosen scope.
- Results show subject, sender, created time, thread id, and a highlighted snippet when using FTS.
- Agents panel shows registered agents for the project with a link to each inbox.
- Quick links to File Reservations and Attachments for the project header.
- Rich search form with filters:
-
/mail/{project}/inbox/{agent}(Inbox for one agent)- Reverse-chronological list with subject, sender, created time, importance badge, thread id.
- Pagination (
?page=N&limit=M).
-
/mail/{project}/message/{id}(Message detail)- Shows subject, sender, created time, importance, recipients (To/Cc/Bcc), thread messages.
- Body rendering:
- If the server pre-converted markdown to HTML, it's sanitized with Bleach (limited tags/attributes, safe CSS via CSSSanitizer) and then displayed.
- Otherwise markdown is rendered client-side with Marked + Prism for code highlighting.
- Attachments are referenced from the message frontmatter (WebP files or inline data URIs).
-
/mail/{project}/search?q=...(Dedicated search page)- Same query syntax as the project overview search, with a token "pill" UI for assembling/removing filters.
-
/mail/{project}/file_reservations(File Reservations list)- Displays active and historical file reservations (exclusive/shared, path pattern, timestamps, released/expired state).
-
/mail/{project}/attachments(Messages with attachments)- Lists messages that contain any attachments, with subject and created time.
-
/mail/unified-inbox(Cross-project activity)- Shows recent messages across all projects with thread counts and sender/recipients.
Human Overseer: Sending Messages to Agents
Sometimes a human operator needs to guide or redirect agents directly, whether to handle an urgent issue, provide clarification, or adjust priorities. The Human Overseer feature provides a web-based message composer that lets humans send high-priority messages to any combination of agents in a project.
Access: Click the prominent "Send Message" button (with the Overseer badge) in the header of any project view (/mail/{project}), or navigate directly to /mail/{project}/overseer/compose.
What Makes Overseer Messages Special
-
Automatic Preamble: Every message includes a formatted preamble that clearly identifies it as coming from a human operator and instructs agents to:
- Pause current work temporarily
- Prioritize the human's request over existing tasks
- Resume original plans afterward (unless modified by the instructions)
-
High Priority: All overseer messages are automatically marked as high importance, ensuring they stand out in agent inboxes.
-
Policy Bypass: Overseer messages bypass normal contact policies, so humans can always reach any agent regardless of their contact settings.
-
Special Sender Identity: Messages come from a special agent named "HumanOverseer" (automatically created per project) with:
- Program:
WebUI - Model:
Human - Contact Policy:
open
- Program:
The Message Preamble
Every overseer message begins with this preamble (automatically prepended):
---
🚨 MESSAGE FROM HUMAN OVERSEER 🚨
This message is from a human operator overseeing this project. Please prioritize
the instructions below over your current tasks.
You should:
1. Temporarily pause your current work
2. Complete the request described below
3. Resume your original plans afterward (unless modified by these instructions)
The human's guidance supersedes all other priorities.
---
[Your message body follows here]
Using the Composer
The composer interface provides:
- Recipient Selection: Checkbox grid of all registered agents (with "Select All" / "Clear" shortcuts)
- Subject Line: Required, shown in agent inboxes
- Message Body: GitHub-flavored Markdown editor with preview
- Thread ID (optional): Continue an existing conversation or start a new one
- Preamble Preview: See exactly how your message will appear to agents
Example Use Cases
Urgent Issue:
Subject: Urgent: Stop migration and revert changes
The database migration in PR #453 is causing data corruption in staging.
Please:
1. Immediately stop any migration-related work
2. Revert commits from the last 2 hours
3. Wait for my review before resuming
I'm investigating the root cause now.
Priority Adjustment:
Subject: New Priority: Security Vulnerability
A critical security vulnerability was just disclosed in our auth library.
Drop your current tasks and:
1. Update `auth-lib` to version 2.4.1 immediately
2. Review all usages in src/auth/
3. Run the full security test suite
4. Report status in thread #892
This takes precedence over the refactoring work.
Clarification:
Subject: Clarification on API design approach
I see you're debating REST vs. GraphQL in thread #234.
Go with REST for now because:
- Our frontend team has more REST experience
- GraphQL adds complexity we don't need yet
- We can always add GraphQL later if needed
Resume the API implementation with REST.
How Agents See Overseer Messages
When agents check their inbox (via fetch_inbox or resource://inbox/{name}), overseer messages appear like any other message but with:
- Sender:
HumanOverseer - Importance:
high(displayed prominently) - Body: Starts with the overseer preamble, followed by the human's message
- Visual cues: In the Web UI, these messages may have special highlighting (future enhancement)
Agents can reply to overseer messages just like any other message, continuing the conversation thread.
Technical Details
- Storage: Overseer messages are stored identically to agent-to-agent messages (Git + SQLite)
- Git History: Fully auditable; message appears in
messages/YYYY/MM/{id}.mdwith commit history - Thread Continuity: Can be part of existing threads or start new ones
- No Authentication Bypass: The overseer compose form still requires proper HTTP server authentication (if enabled)
Design Philosophy
The Human Overseer feature is designed to be:
- Explicit: Agents clearly know when guidance comes from a human vs. another agent
- Respectful: Instructions acknowledge agents have existing work and shouldn't just "drop everything" permanently
- Temporary: Agents are told to resume original plans once the human's request is complete
- Flexible: Humans can override this guidance directly in their message body
This creates a clear hierarchy (human → agents) while maintaining the collaborative, respectful tone of the agent communication system.
Related Projects Discovery
The Projects index (/mail) features an AI-powered discovery system that intelligently suggests which projects should be linked together, such as frontend + backend or related microservices.
How Discovery Works
1. Smart Analysis The system uses multiple signals to identify relationships:
- Pattern matching: Compares project names and paths (e.g., "my-app-frontend" ↔ "my-app-backend")
- AI understanding (when
LLM_ENABLED=true): ReadsREADME.md,AGENTS.md, and other docs to understand each project's purpose and detect natural relationships - Confidence scoring: Ranks suggestions from 0-100% with clear rationales
2. Beautiful Suggestions Related projects appear as polished cards on your dashboard with:
- 🎯 Visual confidence indicators showing match strength
- 💬 AI-generated rationales explaining the relationship
- ✅ Confirm Link - accept the suggestion
- ✖️ Dismiss - hide irrelevant matches
3. Quick Navigation Once confirmed, both projects display interactive badges for instant navigation between related codebases.
Why Suggestions, Not Auto-Linking?
TL;DR: We keep you in control. Discovery helps you find relationships; explicit approvals control who can actually communicate.
Agent Mail uses agent-centric messaging: every message follows explicit permission chains:
Send Message → Find Recipient → Check AgentLink Approval → Deliver
This design ensures:
- Security: No accidental cross-project message delivery
- Transparency: You always know who can talk to whom
- Audit trails: All communication paths are explicitly approved
Why not auto-link with AI? If we let an LLM automatically authorize messaging between projects, we'd be:
- ❌ Bypassing contact policies without human oversight
- ❌ Risking message misdelivery to unintended recipients
- ❌ Creating invisible routing paths that are hard to audit
- ❌ Potentially linking ambiguously-named projects incorrectly
Instead, we give you discovery + control:
- ✅ AI suggests likely relationships (safe, read-only analysis)
- ✅ You confirm what makes sense (one click)
- ✅ Agents still use
request_contact/respond_contactfor actual messaging permissions - ✅ Clear separation: discovery ≠ authorization
The Complete Workflow
1. System suggests: "These projects look related" (AI analysis)
↓
2. You confirm: "Yes, link them" (updates UI badges)
↓
3. Agents request: request_contact(from_agent, to_agent, to_project)
↓
4. You approve: respond_contact(accept=true)
↓
5. Messages flow: Agents can now communicate across projects
Think of it like LinkedIn: The system suggests connections, but only you decide who gets to send messages.
Search syntax (UI)
The UI shares the same parsing as the API's _parse_fts_query:
- Field filters:
subject:login,body:"api key" - Phrase search:
"build plan" - Combine terms:
login AND security(FTS) - Fallback LIKE: scope determines whether subject, body, or both are searched
Prerequisites to see data
The UI reads from the same SQLite + Git artifacts as the MCP tools. To populate content:
- Ensure a project exists (via tool call or CLI):
- Ensure/create project:
ensure_project(human_key)
- Ensure/create project:
- Register one or more agents:
register_agent(project_key, program, model, name?) - Send messages:
send_message(...)(attachments and inline images are supported; images may be converted to WebP).
Once messages exist, visit /mail, click your project, then open an agent inbox or search.
Implementation and dependencies
- Templates live in
src/mcp_agent_mail/templates/and are rendered by Jinja2. - Markdown is converted with
markdown2on the server where possible; HTML is sanitized with Bleach (with CSS sanitizer when available). - Tailwind CSS, Lucide icons, Alpine.js, Marked, and Prism are loaded via CDN in
base.htmlfor a modern look without a frontend build step. - All rendering is server-side; there's no SPA router. Pages degrade cleanly without JavaScript.
Security considerations
- HTML sanitization: Only a conservative set of tags/attributes are allowed; CSS is filtered. Links are limited to http/https/mailto/data.
- Auth: Use bearer token or JWT when exposing beyond localhost. For local dev, enable localhost bypass as noted above.
- Rate limiting (optional): Token-bucket limiter can be enabled; UI GET requests are light and unaffected by POST limits.
Troubleshooting the UI
- Blank page or 401 on localhost: Either unset
HTTP_BEARER_TOKENor setHTTP_ALLOW_LOCALHOST_UNAUTHENTICATED=true. - No projects listed: Create one with
ensure_project. - Empty inbox: Verify recipient names match exactly and messages were sent to that agent.
- Search returns nothing: Try simpler terms or the LIKE fallback (toggle scope/body).
Static Mailbox Export (Share & Distribute Archives)
The share command group exports a project’s mailbox into a portable, read‑only bundle that anyone can review in a browser. It’s designed for auditors, stakeholders, or teammates who need to browse threads, search history, or prove delivery timelines without spinning up the full MCP Agent Mail stack.
Why export to static bundles?
Compliance and audit trails: Deliver immutable snapshots of project communication to auditors or compliance officers. The static bundle includes cryptographic signatures for tamper-evident distribution.
Stakeholder review: Share conversation history with product managers, executives, or external consultants who don't need write access. They can browse messages, search threads, and view attachments in their browser without authentication.
Offline access: Create portable archives for air-gapped environments, disaster recovery backups, or situations where internet connectivity is unreliable.
Long-term archival: Preserve project communication in a format that will remain readable decades from now. Static HTML requires no database server, no runtime dependencies, and survives software obsolescence better than proprietary formats.
Secure distribution: Encrypt bundles with age for confidential projects. Only recipients with the private key can decrypt and view the contents.
What's included in an export
Each bundle contains:
- Self-contained: Everything ships in a single directory (HTML, CSS/JS, SQLite snapshot, attachments). Drop it on a static host or open it locally.
- Rich reader UI: Gmail-style inbox with project filters, search, and full-thread rendering—each message is shown with its metadata and Markdown body, just like in the live web UI.
- Fast search & filters: FTS-backed search and precomputed per-message summaries keep scrolling and filtering responsive even with large archives.
- Verifiable integrity: SHA-256 hashes for every asset plus optional Ed25519 signing make authenticity and tampering checks straightforward.
- Chunk-friendly archives: Large databases can be chunked for httpvfs streaming; a companion
chunks.sha256file lists digests for each chunk so clients can trust streamed blobs without recomputing hashes. - One-click hosting: The interactive wizard can publish straight to GitHub Pages or Cloudflare Pages, or you can serve the bundle locally with the CLI preview command.
Disaster Recovery Archives (archive commands)
Use the archive subcommands when you need a restorable snapshot (not just a read-only share bundle). Each ZIP under ./archived_mailbox_states/ includes:
- A SQLite snapshot processed by the same cleanup pipeline as
share, but using thearchivescrub preset so ack/read state, recipients, attachments, and message bodies remain untouched. - A byte-for-byte copy of the storage Git repo (
STORAGE_ROOT), preserving markdown artifacts, attachments, and hook scripts.
Quick ref
# Save current state (defaults to the lossless preset)
uv run python -m mcp_agent_mail.cli archive save --label nightly
# List available restore points (JSON is handy for scripts)
uv run python -m mcp_agent_mail.cli archive list --json
# Restore after a disaster (backs up any existing DB/storage before overwriting)
uv run python -m mcp_agent_mail.cli archive restore archived_mailbox_states/<file>.zip --force
During restore the CLI:
- Extracts the ZIP into a temp directory.
- Moves any existing
storage.sqlite3, WAL/SHM siblings, andSTORAGE_ROOTinto timestamped.backup-<ts>folders so nothing is lost. - Copies the snapshot back to the configured database path and rebuilds the storage repo from the archive contents.
Every archive writes a metadata.json manifest describing the projects captured, scrub preset, and a friendly reminder of the exact archive restore … command to run later.
Reset safety net
clear-and-reset-everything now offers to create one of these archives before deleting anything. By default it prompts interactively; pass --archive/--no-archive to force a choice, and pair with --force --no-archive for non-interactive automation. When an archive is created successfully, the CLI prints both the path and the restore command so you can undo the reset later.
Mailbox Health: am doctor
The doctor command group provides comprehensive diagnostics and repair capabilities for maintaining mailbox health. It emphasizes data safety—creating backups before any destructive operation and using semi-automatic repair to prevent accidental data loss.
Why am doctor?
Over time, mailbox state can drift:
- Stale locks: Process crashes leave behind
.archive.lockor.commit.lockfiles that block operations - Orphaned records: Agents get deleted but their message recipients remain in the database
- FTS index desync: Full-text search index falls out of sync with actual messages
- Expired file reservations: Reservations expire but aren't cleaned up
- WAL files: SQLite WAL/SHM files accumulate (normal during operation, but worth monitoring)
The doctor commands detect these issues and offer safe, automatic repair.
Diagnostic checks
Run comprehensive diagnostics on your mailbox:
# Check all projects
uv run python -m mcp_agent_mail.cli doctor check
# Check with verbose output
uv run python -m mcp_agent_mail.cli doctor check --verbose
# JSON output for automation
uv run python -m mcp_agent_mail.cli doctor check --json
What it checks:
| Check | Status | Description |
|---|---|---|
| Locks | OK/WARN | Detects stale archive and commit locks from crashed processes |
| Database | OK/ERROR | Runs PRAGMA integrity_check for SQLite corruption |
| Orphaned Records | OK/WARN | Finds message recipients without corresponding agents |
| FTS Index | OK/WARN | Compares message count vs FTS index entries |
| File Reservations | OK/INFO | Counts expired reservations pending cleanup |
| WAL Files | OK/INFO | Reports presence of SQLite WAL/SHM files |
Example output:
MCP Agent Mail Doctor - Diagnostic Report
==================================================
Check Status Details
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Locks OK No stale locks found
Database OK Database integrity check passed
Orphaned OK No orphaned records found
FTS Index OK FTS index synchronized (1,234 messages)
File Res. INFO 4 expired reservation(s) pending cleanup
WAL Files OK No orphan WAL/SHM files
All checks passed!
Semi-automatic repair
The repair command uses a semi-automatic approach:
- Safe repairs (locks, expired reservations) are applied automatically
- Data-affecting repairs (orphan cleanup) require confirmation
- A backup is created before any changes
# Preview what would be repaired (dry-run)
uv run python -m mcp_agent_mail.cli doctor repair --dry-run
# Run repairs with prompts for data changes
uv run python -m mcp_agent_mail.cli doctor repair
# Auto-confirm all repairs (for automation)
uv run python -m mcp_agent_mail.cli doctor repair --yes
# Specify custom backup location
uv run python -m mcp_agent_mail.cli doctor repair --backup-dir /path/to/backups
Repair workflow:
- Create backup — Git bundle + SQLite copy created before any changes
- Safe repairs (auto-applied):
- Heal stale locks (removes orphaned
.archive.lock,.commit.lockfiles) - Release expired file reservations (marks
released_tsin database)
- Heal stale locks (removes orphaned
- Data repairs (require confirmation):
- Delete orphaned message recipients
- Rebuild FTS index (if needed)
Backup management
Doctor creates timestamped backups before repairs. You can also manage backups directly:
# List all available backups
uv run python -m mcp_agent_mail.cli doctor backups
# JSON output for scripting
uv run python -m mcp_agent_mail.cli doctor backups --json
Backup contents:
Each backup includes:
database.sqlite3— Complete SQLite database copydatabase.sqlite3-wal,database.sqlite3-shm— WAL files if presentarchive.bundleor{project}.bundle— Git bundle of the archive repositorymanifest.json— Metadata: when created, why, what's included, restore instructions
Directory structure:
{storage_root}/backups/
2026-01-06T12-30-45_doctor-repair/
manifest.json
database.sqlite3
archive.bundle
Restore from backup
If something goes wrong, restore from any backup:
# Preview what would be restored
uv run python -m mcp_agent_mail.cli doctor restore /path/to/backup --dry-run
# Restore (prompts for confirmation)
uv run python -m mcp_agent_mail.cli doctor restore /path/to/backup
# Skip confirmation prompt
uv run python -m mcp_agent_mail.cli doctor restore /path/to/backup --yes
Restore process:
- Validates backup manifest exists and is readable
- Shows backup metadata (creation time, reason, contents)
- Creates a pre-restore backup of current state (safety net)
- Restores SQLite database from backup
- Restores Git archive from bundle
- Reports any errors encountered
Safety features:
- Current database saved as
*.sqlite3.pre-restorebefore overwrite - Current archive saved as
*.pre-restoredirectory before overwrite - Errors during restore are captured and reported
Best practices
- Run diagnostics regularly:
am doctor checkis fast and non-destructive - Review before repair: Use
--dry-runfirst to see what would change - Keep backups: Don't delete old backups until you've verified the system is healthy
- Automate checks: Include
am doctor check --jsonin your CI/monitoring for early warning
Quick Start: Interactive Deployment Wizard
The easiest way to export and deploy is the interactive wizard, which supports both GitHub Pages and Cloudflare Pages:
# Via CLI (recommended)
uv run python -m mcp_agent_mail.cli share wizard
# Or run the script directly
./scripts/share_to_github_pages.py
What the wizard does
The wizard provides a fully automated end-to-end deployment experience:
- Session resumption: Detects interrupted sessions and offers to resume exactly where you left off, avoiding re-export
- Configuration management: Remembers your last settings and offers to reuse them, saving time on repeated exports
- Deployment target selection: Choose between GitHub Pages, Cloudflare Pages, or local export
- Automatic CLI installation: Detects and installs missing tools (
ghfor GitHub,wranglerfor Cloudflare) - Guided authentication: Step-by-step browser login flows for GitHub and Cloudflare
- Smart project selection:
- Shows all available projects in a formatted table
- Supports multiple selection modes:
all, single number (1), lists (1,3,5), or ranges (1-3,2-5,8) - Remembers your previous selection for quick re-export
- Redaction configuration: Choose between
standard(scrub secrets like API keys/tokens, keep agent names) orstrict(redact all message bodies) - Cryptographic signing: Optional Ed25519 signing with automatic key generation or reuse of existing keys
- Pre-flight validation: Checks that GitHub repo names are available before starting the export
- Deployment summary: Shows what will be deployed (project count, bundle size, target, signing status) and asks for confirmation
- Export and preview: Exports the bundle and launches an interactive preview server with automatic port detection (tries 9000-9100)
- Interactive preview controls:
- Press 'r' to force browser refresh (manual cache bust)
- Press 'd' to skip preview and deploy immediately
- Press 'q' to quit preview server
- Automatic viewer asset refresh: Always ensures latest HTML/JS/CSS from source tree are used, even when reusing bundles
- Real-time deployment: Streams git and deployment output in real-time so you can follow the progress
- Automatic deployment: Creates repos, enables Pages, pushes code, and gives you the live URL
Session resumption (new in latest version)
If you interrupt the wizard (close terminal, Ctrl+C during preview, etc.), it saves your progress to ~/.mcp-agent-mail/wizard-session/. When you run the wizard again:
Incomplete session detected
Projects: 3 selected
Stage: preview
Workspace: ~/.mcp-agent-mail/wizard-session/bundle
Resume where you left off? (Y/n):
What gets saved:
- Selected projects and scrub preset
- Deployment configuration (target, repo name, etc.)
- Signing key preferences and paths
- Exported bundle (in session workspace)
- Current stage (preview, deploy)
Resume scenarios:
- Closed terminal during preview: Resume → Skip re-export → Launch preview immediately
- Changed your mind after export: Resume → "Reuse bundle?" → Preview or re-export
- Want to deploy later: Resume → Press 'd' in preview → Deploy without re-exporting
- Made viewer code changes: Resume → Assets auto-refresh from source tree
After successful deployment, the session state is automatically cleared. Sessions also clear if they become invalid (workspace deleted, projects removed, etc.).
Configuration persistence
The wizard saves your configuration to ~/.mcp-agent-mail/wizard-config.json after each successful deployment. On subsequent runs, it will show:
Previous Configuration Found
Projects: 3 selected
Redaction: standard
Target: github-new
Use these settings again? (Y/n):
This allows rapid re-deployment with the same settings. The saved configuration includes:
- Selected project indices (validates against current project list)
- Redaction preset
- Deployment target and parameters (repo name, privacy, project name)
- Signing preferences (whether to sign, whether to generate new key)
- Last used signing key path (offered as default when not generating new key)
Configuration is project-agnostic: if you add or remove projects, the wizard validates saved indices and prompts for re-selection if needed.
Difference between session and config:
- Session state (
wizard-session/): Temporary, for resuming interrupted runs, includes exported bundle - Config file (
wizard-config.json): Persistent, for "use last settings" across fresh runs, no bundle
Multi-project selection
The project selector supports flexible selection syntax:
Available Projects:
# Slug Path
1 backend-abc123 /abs/path/backend
2 frontend-xyz789 /abs/path/frontend
3 infra-def456 /abs/path/infra
4 scripts-ghi789 /abs/path/scripts
Select projects to export (e.g., 'all', '1,3,5', or '1-3'):
Selection modes:
all: Export all projects (default)1: Export project #1 only1,3,5: Export projects #1, #3, and #51-3: Export projects #1, #2, and #3 (inclusive range)2-4,7: Export projects #2, #3, #4, and #7 (combined range and list)
Invalid selections (out of range, malformed) are rejected with helpful error messages and the wizard prompts again.
Dynamic port allocation
The preview server automatically detects an available port in the range 9000-9100 instead of failing if port 9000 is in use. The actual port is displayed:
Launching preview server...
Using port 9001 (Ctrl+C to stop server)
Waiting for server to start...
✓ Server ready, opening browser at http://127.0.0.1:9001
This prevents port conflicts when multiple previews are running or when port 9000 is used by other services.
Deployment summary panel
Before starting the export, the wizard shows a comprehensive summary:
═══ Deployment Summary ═══
Projects: 3 selected
Bundle size: ~32 MB
Redaction: standard
Target: GitHub Pages
Repository: mailbox-viewer-2024
Visibility: Private
Signing: Enabled (Ed25519)
Proceed with export and deployment? (Y/n):
This gives you a final chance to review all settings and cancel if needed. The bundle size is estimated based on ~10 MB per project plus ~2 MB for static assets.
Real-time deployment streaming
Git operations and Cloudflare deployments stream output in real-time so you can see exactly what's happening:
Initializing git repository and pushing...
Initializing repository...
Initialized empty Git repository in /tmp/mailbox-preview-abc123/.git/
✓ Initializing repository complete
Adding files...
✓ Adding files complete
Creating commit...
[main (root-commit) 1a2b3c4] Initial mailbox export
425 files changed, 123456 insertions(+)
✓ Creating commit complete
Pushing to GitHub...
Enumerating objects: 430, done.
Counting objects: 100% (430/430), done.
Delta compression using up to 8 threads
Compressing objects: 100% (425/425), done.
Writing objects: 100% (430/430), 12.34 MiB | 5.67 MiB/s, done.
✓ Pushing to GitHub complete
✓ Successfully pushed to owner/mailbox-viewer-2024
This provides transparency and helps diagnose issues if deployment fails.
Platform-specific details
For GitHub Pages:
- Wizard detects your package manager (brew/apt/dnf) and offers automated installation of
ghCLI - For apt/dnf, shows complete manual installation instructions (including repo setup) since automation requires sudo
- Runs
gh auth logininteractively to authenticate via browser - Creates new repository with your specified name and visibility (public/private)
- Initializes git, commits, and pushes with streaming output
- Enables GitHub Pages automatically via the GitHub API
- Provides the GitHub Pages URL (may take 1-2 minutes to become live)
For Cloudflare Pages:
- Detects npm and offers automated installation of
wranglerCLI - Runs
wrangler logininteractively to authenticate via browser - Deploys directly to Cloudflare's global CDN (no git repository needed)
- Streams wrangler output in real-time
- Provides the
.pages.devURL immediately (site is live instantly) - Benefits: instant deployment, 275+ global locations, automatic HTTPS, unlimited requests on free tier
For local export:
- Saves bundle to specified directory
- No CLI installation or authentication required
- Suitable for manual deployment to custom hosting or inspection
Error handling and recovery
The wizard includes comprehensive error handling:
- Pre-flight validation: Checks GitHub repo availability before starting export to avoid conflicts
- Port conflict resolution: Automatically finds an available port for preview server
- Invalid selection handling: Validates project selections and prompts for correction
- CLI installation failures: Shows manual installation instructions if automatic installation fails
- Git operation failures: Each git step is validated; stops on first failure with clear error message
- Deployment failures: Distinguishes between repo creation, push, and Pages enablement failures
If deployment fails after export, the bundle remains in the temp directory and can be deployed manually using the git commands shown in the manual deployment section below.
The wizard handles all operations automatically. For manual control or advanced options, see the detailed workflows below.
Basic export workflow (manual)
1. Export a bundle
# Export all projects to a directory
uv run python -m mcp_agent_mail.cli share export --output ./my-bundle
# Export specific projects only
uv run python -m mcp_agent_mail.cli share export \
--output ./my-bundle \
--project backend-abc123 \
--project frontend-xyz789
# Export with Ed25519 signing for tamper-evident distribution
uv run python -m mcp_agent_mail.cli share export \
--output ./my-bundle \
--signing-key ./keys/signing.key \
--signing-public-out ./keys/signing.pub
# Export and encrypt with age for secure distribution
uv run python -m mcp_agent_mail.cli share export \
--output ./my-bundle \
--age-recipient age1abc...xyz \
--age-recipient age1def...uvw
The export process:
- Creates a snapshot of the SQLite database (read-only, no WAL/SHM files)
- Copies message bodies, attachments, and metadata into the bundle structure
- Applies redaction rules based on the scrub preset (default:
standard) - Generates
manifest.jsonwith SHA-256 hashes for all assets - Optionally signs the manifest with Ed25519 (produces
manifest.sig.json) - Packages everything into a ZIP archive (optional, enabled by default)
- If chunking is enabled, writes the segmented database plus a
chunks.sha256manifest so streamed pages can be verified cheaply - Optionally encrypts the ZIP with age (produces
bundle.zip.age)
Refresh an existing bundle
Once you have published a bundle you can refresh it in place without re-running the full wizard. Every export records the settings that were used (projects, scrub preset, attachment thresholds, chunking config) inside manifest.json. The new share update command reads those defaults, regenerates the SQLite snapshot and viewer assets in a temporary directory, and then replaces the bundle atomically—removing obsolete chunked files or attachments along the way.
# Refresh bundle using the originally recorded settings
uv run python -m mcp_agent_mail.cli share update ./my-bundle
# Override one or more export options while updating
uv run python -m mcp_agent_mail.cli share update ./my-bundle \
--project backend-abc123 \
--inline-threshold 16384 \
--chunk-threshold 104857600
# Re-sign and package the refreshed bundle
uv run python -m mcp_agent_mail.cli share update ./my-bundle \
--zip \
--signing-key ./keys/signing.key
When chunking was enabled previously but the refreshed snapshot no longer needs it, share update cleans up the chunks/ directory, chunks.sha256, and mailbox.sqlite3.config.json automatically, ensuring the bundle tree matches the new manifest. You can still tweak any setting at update time; overrides are written back into the export_config section of manifest.json for the next refresh.
2. Preview locally
# Serve the bundle on localhost:9000
uv run python -m mcp_agent_mail.cli share preview ./my-bundle
# Custom port and auto-open browser
uv run python -m mcp_agent_mail.cli share preview ./my-bundle \
--port 8080 \
--open-browser
This launches a lightweight HTTP server that serves the static files. Open http://127.0.0.1:9000/viewer/ in your browser to explore the archive.
Interactive preview controls:
- 'r': Force browser reload (bumps manual cache-bust token, triggers viewer refresh)
- 'd': Request deployment (exits with code 42; wizard detects and proceeds to deploy)
- 'q': Quit preview server
- Ctrl+C: Stop preview server
The preview server automatically refreshes viewer assets from the source tree if available, ensuring you always see the latest HTML/JS/CSS during development.
3. Verify integrity
# Verify SRI hashes and signature
uv run python -m mcp_agent_mail.cli share verify ./my-bundle
# Verify with explicit public key (overrides manifest.sig.json)
uv run python -m mcp_agent_mail.cli share verify ./my-bundle \
--public-key AAAA...base64...
Verification checks:
- SHA-256 hashes for all vendor libraries (Marked.js, DOMPurify, SQL.js)
- SHA-256 hashes for the SQLite database and attachments
- Ed25519 signature over the canonical manifest (if present)
4. Decrypt (if encrypted)
# Decrypt with age identity file (private key)
uv run python -m mcp_agent_mail.cli share decrypt bundle.zip.age \
--identity ~/.age/key.txt
# Decrypt with passphrase (interactive prompt)
uv run python -m mcp_agent_mail.cli share decrypt bundle.zip.age \
--passphrase
# Specify custom output path
uv run python -m mcp_agent_mail.cli share decrypt bundle.zip.age \
--output ./decrypted-bundle.zip \
--identity ~/.age/key.txt
After decryption, unzip the archive and use share preview to view it.
Export options reference
| Option | Type | Default | Description |
|---|---|---|---|
--output, -o | Path | (required) | Directory where the static bundle will be written |
--project, -p | List | All projects | Limit export to specific project slugs or human keys (repeatable) |
--inline-threshold | Bytes | 65536 (64KB) | Inline attachments smaller than this as base64 data URIs |
--detach-threshold | Bytes | 26214400 (25MB) | Mark attachments larger than this as external (not bundled) |
--scrub-preset | String | standard | Redaction preset: standard or strict (see Redaction presets section) |
--chunk-threshold | Bytes | 20971520 (20MB) | Split SQLite database into chunks if it exceeds this size |
--chunk-size | Bytes | 4194304 (4MB) | Chunk size when splitting large databases |
--dry-run | Flag | false | Generate security summary and preview without writing files |
--zip / --no-zip | Flag | true | Package the bundle into a ZIP archive |
--signing-key | Path | None | Path to Ed25519 signing key (32-byte raw seed) |
--signing-public-out | Path | None | Write the Ed25519 public key to this file after signing |
--age-recipient | String | None | age public key for encryption (repeatable for multiple recipients) |
--interactive, -i | Flag | false | Launch interactive wizard (prints guidance; full wizard TBD) |
Security features
XSS protection (DOMPurify + CSP)
Message bodies are rendered using a defense-in-depth pipeline:
- Marked.js parses GitHub-Flavored Markdown into HTML
- DOMPurify sanitizes the HTML, removing dangerous tags and attributes
- Content Security Policy restricts script sources, blocks inline event handlers, and limits network access
This prevents malicious content in message bodies from executing JavaScript or exfiltrating data.
CSP configuration notes:
script-src: Allows self, CDNs (Tailwind, Alpine.js), and'unsafe-eval'(required for SQL.js WebAssembly)connect-src: Allows*(all origins) to support preview mode polling and flexible deployment scenariosstyle-src: Allows self, inline styles (for Tailwind), and font CDNs- Trusted Types removed for browser compatibility (Firefox, Safari don't support it yet)
Cryptographic signing (Ed25519)
When you provide a signing key, the export process:
- Generates a canonical JSON representation of
manifest.json - Signs it with Ed25519 (fast, 64-byte signatures, 128-bit security)
- Writes the signature and public key to
manifest.sig.json
Recipients can verify the signature using share verify to ensure:
- The bundle hasn't been modified since signing
- The bundle was created by someone with the private key
- All assets match their declared SHA-256 hashes
Requirements and fallback:
- Requires PyNaCl >= 1.6.0 (installed automatically with this package)
- If PyNaCl is unavailable or signing fails, export gracefully falls back to unsigned mode
- Wizard reuses existing signing keys by default (no re-generation unless requested)
- Private keys are automatically excluded from git via
.gitignore(signing-*.key pattern)
Encryption (age)
The age encryption tool (https://age-encryption.org/) provides modern, secure file encryption. When you provide recipient public keys, the export process encrypts the final ZIP archive. Only holders of the corresponding private keys can decrypt it.
Generate keys with:
# Install age (example for macOS)
brew install age
# Generate a new key pair
age-keygen -o key.txt
# Public key is printed to stdout; share it with exporters
# Private key is saved to key.txt; keep it secret
Redaction presets
The export pipeline supports configurable scrubbing to remove sensitive data:
-
standard: Clears acknowledgment/read state, removes file reservations and agent links, scrubs secrets (GitHub tokens, Slack tokens, OpenAI keys, bearer tokens, JWTs) from message bodies and attachment metadata. Retains agent names (which are already meaningless pseudonyms like "BlueMountain"), full message bodies, and attachments. -
strict: All standard redactions plus replaces entire message bodies with "[Message body redacted]" placeholder and removes all attachments from the bundle.
All presets apply redaction to message subjects, bodies, and attachment metadata before the bundle is written.
Static viewer features
The bundled HTML viewer provides:
Dashboard layout:
- Gmail-style three-pane interface: Projects sidebar, message list (center), and detail pane (right)
- Bundle metadata header: Shows bundle creation time, export settings, and scrubbing preset
- Summary panels: Side-by-side panels displaying projects included, attachment statistics, and redaction summary
- Message list: Virtual-scrolled message list with sender, subject, snippet, and importance badges
- Raw manifest viewer: Collapsible JSON display of the complete manifest for verification
Advanced boolean search (new): Powered by SQLite FTS5 with LIKE fallback, supports complex queries:
- Boolean operators:
(auth OR login) AND NOT admin - Quoted phrases:
"build plan"(exact match) - Parentheses: Control precedence like
(A OR B) AND (C OR D) - Operator precedence: NOT > AND > OR (e.g.,
A OR B AND C=A OR (B AND C)) - Automatic debouncing: 140ms delay avoids hammering the database on every keystroke
- Performance: FTS5 search is 10-100x faster than LIKE on large datasets
Lazy message loading (performance optimization):
- Initial load fetches only 280-character snippets for all messages (3-6x faster)
- Full message body loaded on-demand when you click a message
- Dramatically reduces memory usage and initial load time
Virtual scrolling (new): Clusterize.js-powered virtual list rendering:
- Smoothly handles 100,000+ messages without slowdown
- Only ~30 DOM nodes exist at any time (visible rows + buffers)
- Maintains native scrollbar feel with keyboard navigation
Markdown rendering: Message bodies are rendered with GitHub-Flavored Markdown, supporting code blocks, tables, task lists, and inline images.
Opportunistic OPFS caching: The SQLite database is cached in Origin Private File System (OPFS) in the background:
- First load: Downloads from network, caches to OPFS during idle time
- Subsequent loads: Instant from OPFS (even faster than IndexedDB)
- Automatic cache key validation prevents stale data
Dark mode: Toggle between light and dark themes with localStorage persistence. Dark mode state is managed by the main viewer controller for consistency.
Attachment preview: Inline images render directly in message bodies. External attachments show file size and download links.
Message detail view: Click any message in the list to load its full body (lazy), view metadata (sender, recipients, timestamp, importance), and browse attachments.
No server required: After the initial HTTP serving (which can be a static file host like S3, GitHub Pages, or Netlify), all functionality runs client-side. No backend, no API calls, no authentication.
Browser compatibility: Works in all modern browsers (Chrome, Firefox, Safari, Edge) with graceful fallbacks for missing features (OPFS, FTS5).
Deployment options
Option 1: GitHub Pages (automated via wizard)
# Use the wizard for fully automated deployment
uv run python -m mcp_agent_mail.cli share wizard
# Select: GitHub Pages → provide repo name → wizard handles everything
Or manually:
# Export and unzip
uv run python -m mcp_agent_mail.cli share export --output ./bundle --no-zip
cd bundle
# Initialize git and push to GitHub
git init
git add .
git commit -m "Initial export"
git remote add origin git@github.com:yourorg/project-archive.git
git push -u origin main
# Enable GitHub Pages in repo settings (source: main branch, root directory)
Option 2: Cloudflare Pages (automated via wizard)
# Use the wizard for instant global CDN deployment
uv run python -m mcp_agent_mail.cli share wizard
# Select: Cloudflare Pages → provide project name → wizard deploys directly
Or manually with wrangler CLI:
# Export and deploy
uv run python -m mcp_agent_mail.cli share export --output ./bundle --no-zip
npx wrangler pages deploy ./bundle --project-name=project-archive
# Your site is live at: https://project-archive.pages.dev
Benefits of Cloudflare Pages:
- Instant deployment (no git repo required)
- Global CDN with 275+ locations
- Automatic HTTPS and DDoS protection
- Zero-downtime updates
- Generous free tier (500 builds/month, unlimited requests)
Option 3: S3 + CloudFront
# Export and unzip
uv run python -m mcp_agent_mail.cli share export --output ./bundle --no-zip
# Upload to S3
aws s3 sync ./bundle s3://your-bucket/archives/project-2024/ --acl public-read
# Access via CloudFront
# https://d123abc.cloudfront.net/archives/project-2024/
Option 4: Nginx static site
server {
listen 443 ssl;
server_name archives.example.com;
ssl_certificate /etc/letsencrypt/live/archives.example.com/fullchain.pem;
ssl_certificate_key /etc/letsencrypt/live/archives.example.com/privkey.pem;
root /var/www/archives/project-2024;
index index.html;
# Enable gzip for efficient transfer
gzip on;
gzip_types text/html text/css application/javascript application/json application/wasm;
# Cache static assets
location ~* \.(js|css|wasm|png|jpg|webp)$ {
expires 1y;
add_header Cache-Control "public, immutable";
}
# CSP headers are already in index.html meta tag
# Add HTTPS-only and frame protection
add_header Strict-Transport-Security "max-age=31536000; includeSubDomains" always;
add_header X-Frame-Options "DENY" always;
add_header X-Content-Type-Options "nosniff" always;
}
Option 5: Encrypted distribution via file sharing
For confidential archives:
# Export with age encryption
uv run python -m mcp_agent_mail.cli share export \
--output ./bundle \
--signing-key ./signing.key \
--age-recipient age1auditor... \
--age-recipient age1manager...
# This produces bundle.zip.age
# Upload to Dropbox, Google Drive, or send via secure file transfer
# Recipients decrypt locally
uv run python -m mcp_agent_mail.cli share decrypt bundle.zip.age \
--identity ~/.age/key.txt
# Verify integrity
unzip bundle.zip
uv run python -m mcp_agent_mail.cli share verify ./bundle
# Preview locally
uv run python -m mcp_agent_mail.cli share preview ./bundle
Example workflows
Quarterly audit package
# Export Q4 2024 communications for audit
uv run python -m mcp_agent_mail.cli share export \
--output ./audit-q4-2024 \
--scrub-preset strict \
--signing-key ./audit-signing.key \
--signing-public-out ./audit-signing.pub \
--age-recipient age1auditor@firm.example
# Produces: audit-q4-2024.zip.age
# Send to auditor with audit-signing.pub
# Auditor verifies:
age -d -i auditor-key.txt audit-q4-2024.zip.age > audit-q4-2024.zip
unzip audit-q4-2024.zip
python -m mcp_agent_mail.cli share verify ./audit-q4-2024 \
--public-key $(cat audit-signing.pub)
Executive summary for stakeholders
# Export high-importance threads only
# (filter in UI after export, or use SQL to create filtered snapshot)
uv run python -m mcp_agent_mail.cli share export \
--output ./exec-summary \
--project backend-prod \
--scrub-preset standard
# Host on internal web server
cp -r ./exec-summary /var/www/exec-archives/2024-12/
# Share link: https://internal.example.com/exec-archives/2024-12/
Disaster recovery backup
# Monthly encrypted backup
uv run python -m mcp_agent_mail.cli share export \
--output ./backup-$(date +%Y-%m) \
--scrub-preset none \
--signing-key ./dr-signing.key \
--age-recipient age1dr@company.example
# Store in off-site backup system
aws s3 cp backup-2024-12.zip.age s3://dr-backups/mcp-mail/ \
--storage-class GLACIER_IR
# Restore procedure documented in runbook
Troubleshooting exports
Export fails with "Database locked"
The export takes a snapshot using SQLite's Online Backup API. If the server is actively writing, wait a few seconds and retry. For large databases, consider temporarily stopping the server during export.
Bundle size is too large
Use --detach-threshold to mark large attachments as external references. These won't be included in the bundle but will show file metadata in the viewer.
# Bundle files under 1MB, mark larger files as external
uv run python -m mcp_agent_mail.cli share export \
--output ./bundle \
--detach-threshold 1048576
Alternatively, filter to specific projects with --project.
Encrypted bundle won't decrypt
Verify you're using the correct identity file:
# Get your public key from your identity file
age-keygen -y identity.txt
# Ensure this public key was included in the --age-recipient values during export
# If you have multiple identity files, try each one
age -d -i identity.txt bundle.zip.age > bundle.zip
Signature verification fails
Signature verification requires:
- The original
manifest.json(unmodified) - The
manifest.sig.jsonfile (contains signature and public key) - All assets referenced in the manifest with matching SHA-256 hashes
If verification fails, the bundle may have been tampered with or corrupted during transfer. Re-export and re-transfer.
Viewer shows blank page or errors
Check browser console for errors. Common issues:
- OPFS not supported: Older browsers may not support Origin Private File System. The viewer will fall back to in-memory mode (slower).
- Database too large: Browsers limit in-memory database size to ~1-2GB. Use chunking (
--chunk-threshold) for very large archives. - CSP violations: If hosting the bundle, ensure the web server doesn't add conflicting CSP headers. The viewer's CSP is defined in
index.htmland should not be overridden.
On-disk layout (per project)
<store>/projects/<slug>/
agents/<AgentName>/profile.json
agents/<AgentName>/inbox/YYYY/MM/<msg-id>.md
agents/<AgentName>/outbox/YYYY/MM/<msg-id>.md
messages/YYYY/MM/<msg-id>.md
messages/threads/<thread-id>.md # optional human digest maintained by the server
file_reservations/<sha1-of-path>.json
attachments/<xx>/<sha1>.webp
Message file format
Messages are GitHub-Flavored Markdown with JSON frontmatter (fenced by ---json). Attachments are either WebP files referenced by relative path or inline base64 WebP data URIs.
---json
{
"id": 1234,
"thread_id": "TKT-123",
"project": "/abs/path/backend",
"project_slug": "backend-abc123",
"from": "GreenCastle",
"to": ["BlueLake"],
"cc": [],
"created": "2025-10-23T15:22:14Z",
"importance": "high",
"ack_required": true,
"attachments": [
{"type": "file", "media_type": "image/webp", "path": "projects/backend-abc123/attachments/2a/2a6f.../diagram.webp"}
]
}
---
# Build plan for /api/users routes
... body markdown ...
Data model (SQLite)
projects(id, human_key, slug, created_at)agents(id, project_id, name, program, model, task_description, inception_ts, last_active_ts, attachments_policy, contact_policy)messages(id, project_id, sender_id, thread_id, subject, body_md, created_ts, importance, ack_required, attachments)message_recipients(message_id, agent_id, kind, read_ts, ack_ts)file_reservations(id, project_id, agent_id, path_pattern, exclusive, reason, created_ts, expires_ts, released_ts)agent_links(id, a_project_id, a_agent_id, b_project_id, b_agent_id, status, reason, created_ts, updated_ts, expires_ts)project_sibling_suggestions(id, project_a_id, project_b_id, score, status, rationale, created_ts, evaluated_ts, confirmed_ts, dismissed_ts)fts_messages(message_id UNINDEXED, subject, body)+ triggers for incremental updates
Concurrency and lifecycle
- One request/task = one isolated operation
- Archive writes are guarded by a per-project
.archive.lockunderprojects/<slug>/ - Git index/commit operations are serialized across the shared archive repo by a repo-level
.commit.lock - DB operations are short-lived and scoped to each tool call; FTS triggers keep the search index current
- Artifacts are written first, then committed as a cohesive unit with a descriptive message
- Attachments are content-addressed (sha1) to avoid duplication
How it works (key flows)
- Create an identity
register_agent(project_key, program, model, name?, task_description?)→ creates/updates a named identity, persists profile to Git, and commits.
- Send a message
send_message(project_key, sender_name, to[], subject, body_md, cc?, bcc?, attachment_paths?, convert_images?, importance?, ack_required?, thread_id?, auto_contact_if_blocked?)- Writes a canonical message under
messages/YYYY/MM/, an outbox copy for the sender, and inbox copies for each recipient; commits all artifacts. - Optionally converts images (local paths or data URIs) to WebP and embeds small ones inline.
sequenceDiagram
participant Agent
participant Server
participant DB
participant Git
Agent->>Server: call send_message
Server->>DB: insert message and recipients
DB-->>Server: ok
Server->>Git: write canonical markdown
Server->>Git: write outbox copy
Server->>Git: write inbox copies
Server->>Git: commit
Server-->>Agent: result
- Check inbox
fetch_inbox(project_key, agent_name, since_ts?, urgent_only?, include_bodies?, limit?)returns recent messages, preserving thread_id where available.acknowledge_message(project_key, agent_name, message_id)marks acknowledgements.
- Avoid conflicts with file reservations (leases)
file_reservation_paths(project_key, agent_name, paths[], ttl_seconds, exclusive, reason)records an advisory lease in DB and writes JSON reservation artifacts in Git; conflicts are reported if overlapping active exclusives exist (reservations are still granted; conflicts are returned alongside grants).release_file_reservations(project_key, agent_name, paths? | file_reservation_ids?)releases active leases (all if none specified). JSON artifacts remain for audit history.- Optional: install a pre-commit hook in your code repo that blocks commits conflicting with other agents' active exclusive file reservations.
sequenceDiagram
participant Agent
participant Server
participant DB
participant Git
Agent->>Server: call file_reservation_paths
Server->>DB: expire old leases and check overlaps
DB-->>Server: conflicts or grants
Server->>DB: insert file reservation rows
Server->>Git: write file reservation JSON files
Server->>Git: commit
Server-->>Agent: granted paths and any conflicts
- Search & summarize
search_messages(project_key, query, limit?)uses FTS5 over subject and body.summarize_thread(project_key, thread_id, include_examples?)extracts key points, actions, and participants from the thread.reply_message(project_key, message_id, sender_name, body_md, ...)creates a subject-prefixed reply, preserving or creating a thread.
Semantics & invariants
- Identity
- Names are memorable adjective+noun and unique per project;
name_hintis sanitized (alnum) and used if available whoisreturns the stored profile;list_agentscan filter by recent activitylast_active_tsis bumped on relevant interactions (messages, inbox checks, etc.)
- Names are memorable adjective+noun and unique per project;
- Threads
- Replies inherit
thread_idfrom the original; if missing, the reply setsthread_idto the original message id - Subject lines are prefixed (e.g.,
Re:) for readability in mailboxes
- Replies inherit
- Attachments
- Image references (file path or data URI) are converted to WebP; small images embed inline when policy allows
- Non-absolute paths resolve relative to the project repo root
- Stored under
attachments/<xx>/<sha1>.webpand referenced by relative path in frontmatter
- File Reservations
- TTL-based; exclusive means "please don't modify overlapping surfaces" for others until expiry or release
- Conflict detection is per exact path pattern; shared reservations can coexist, exclusive conflicts are surfaced
- JSON artifacts remain in Git for audit even after release (DB tracks release_ts)
- Search
- External-content FTS virtual table and triggers index subject/body on insert/update/delete
- Queries are constrained to the project id and ordered by
created_ts DESC
Contact model and "consent-lite" messaging
Goal: make coordination "just work" without spam across unrelated agents. The server enforces per-project isolation by default and adds an optional consent layer within a project so agents only contact relevant peers.
Isolation by project
- All tools require a
project_key. Agents only see messages addressed to them within that project. - An agent working in Project A is invisible to agents in Project B unless explicit cross-project contact is established (see below). This avoids distraction between unrelated repositories.
Policies (per agent)
open: accept any targeted messages in the project.auto(default): allow messages when there is obvious shared context (e.g., same thread participants; recent overlapping active file reservations; recent prior direct contact within a TTL); otherwise requires a contact request.contacts_only: require an approved contact link first.block_all: reject all new contacts (errors with CONTACT_BLOCKED).
Use set_contact_policy(project_key, agent_name, policy) to update.
Request/approve contact
request_contact(project_key, from_agent, to_agent, reason?, ttl_seconds?)creates or refreshes a pending link and sends a small ack_required "intro" message to the recipient.respond_contact(project_key, to_agent, from_agent, accept, ttl_seconds?)approves or denies; approval grants messaging until expiry.list_contacts(project_key, agent_name)surfaces current links.
Auto-allow heuristics (no explicit request required)
- Same thread: replies or messages to thread participants are allowed.
- Recent overlapping file reservations: if sender and recipient hold active file reservations in the project, messaging is allowed.
- Recent prior contact: a sliding TTL allows follow-ups between the same pair.
These heuristics minimize friction while preventing cold spam.
Cross-project coordination (frontend vs backend repos)
When two repos represent the same underlying project (e.g., frontend and backend), you have two options:
-
Use the same
project_keyacross both workspaces. Agents in both repos operate under one project namespace and benefit from full inbox/outbox coordination automatically. -
Keep separate
project_keys and establish explicit contact:- In
backend, agentGreenCastlecalls:request_contact(project_key="/abs/path/backend", from_agent="GreenCastle", to_agent="BlueLake", reason="API contract changes")
- In
frontend,BlueLakecalls:respond_contact(project_key="/abs/path/backend", to_agent="BlueLake", from_agent="GreenCastle", accept=true)
- After approval, messages can be exchanged; in default
autopolicy the server allows follow-up threads/reservation-based coordination without re-requesting.
- In
Important: You can also create reciprocal links or set open policy for trusted pairs. The consent layer is on by default (CONTACT_ENFORCEMENT_ENABLED=true) but is designed to be non-blocking in obvious collaboration contexts.
Resource layer (read-only URIs)
Expose common reads as resources that clients can fetch. See API Quick Reference → Resources for the full list and parameters.
Example (conceptual) resource read:
{
"method": "resources/read",
"params": {"uri": "resource://inbox/BlueLake?project=/abs/path/backend&limit=20"}
}
sequenceDiagram
participant Client
participant Server
participant DB
Client->>Server: read inbox resource
Server->>DB: select messages for agent
DB-->>Server: rows
Server-->>Client: inbox data
File Reservations and the optional pre-commit guard
- Guard status and pre-push
- Print guard status:
mcp-agent-mail guard status /path/to/repo
- Install both guards (pre-commit + pre-push):
mcp-agent-mail guard install <project_key> <repo_path> --prepush
- Pre-commit honors
WORKTREES_ENABLEDandAGENT_MAIL_GUARD_MODE(warnadvisory). - Pre-push enumerates to-be-pushed commits (
rev-list) and usesdiff-treewith--no-ext-diff. - Composition-safe install (chain-runner):
- A Python chain-runner is written to
.git/hooks/pre-commitand.git/hooks/pre-push. - It executes
hooks.d/<hook>/*in lexical order, then<hook>.origif present (existing hooks are preserved, not overwritten). - Agent Mail installs its guard as
hooks.d/pre-commit/50-agent-mail.pyandhooks.d/pre-push/50-agent-mail.py. - Windows shims (
pre-commit.cmd/.ps1,pre-push.cmd/.ps1) are written to invoke the Python chain-runner.
- A Python chain-runner is written to
- Matching and safety details:
- Renames/moves are handled: both the old and new names are checked (
git diff --cached --name-status -M -z). - NUL-safe end-to-end: paths are collected and forwarded as NUL-delimited to avoid ambiguity.
- Git-native matching: reservations are checked using Git wildmatch pathspec semantics against repo-root relative paths;
core.ignorecaseis honored. - Emergency bypass (use sparingly): set
AGENT_MAIL_BYPASS=1, or use native Git--no-verify. Inwarnmode the guard never blocks.
- Renames/moves are handled: both the old and new names are checked (
- Print guard status:
Git-based project identity (opt-in)
-
Gate:
WORKTREES_ENABLED=1orGIT_IDENTITY_ENABLED=1enables git-based identity features. Default off. -
Identity modes (default
dir):dir,git-remote,git-toplevel,git-common-dir. -
Inspect identity for a path:
- Resource (MCP):
resource://identity/{/abs/path}(available whenWORKTREES_ENABLED=1) - CLI (diagnostics):
mcp-agent-mail mail status /abs/path
- Resource (MCP):
-
Precedence (when gate is on):
- Committed marker
.agent-mail-project-id(recommended) - Discovery YAML
.agent-mail.yamlwithproject_uid: - Private marker under Git common dir
.git/agent-mail/project-id - Remote fingerprint: normalized
originURL + default branch git-common-dirhash; else dir hash
- Committed marker
-
Migration helpers:
- Write committed marker:
mcp-agent-mail projects mark-identity . --commit - Scaffold discovery file:
mcp-agent-mail projects discovery-init . --product <product_uid>
- Write committed marker:
Example identity payload (resource):
{
"project_uid": "c5b2c86b-7c36-4de6-9a0a-2c4e1c3a1c4a",
"slug": "repo-a1b2c3d4e5",
"identity_mode_used": "git-remote",
"canonical_path": "github.com/owner/repo",
"human_key": "/abs/worktree/path",
"repo_root": "/abs/repo",
"git_common_dir": "/abs/repo/.git",
"branch": "feature/x",
"worktree_name": "repo-wt-x",
"core_ignorecase": true,
"normalized_remote": "github.com/owner/repo"
}
Adopt/Merge legacy projects (optional)
Consolidate legacy per-worktree projects into a canonical one (safe, explicit, and auditable).
- Plan the merge (no changes):
mcp-agent-mail projects adopt <from> <to> --dry-run
- Apply the merge (moves artifacts and re-keys DB rows):
mcp-agent-mail projects adopt <from> <to> --apply
- Safeguards and behavior:
- Requires both projects be in the same repository (validated via
git-common-dir). - Moves archived Git artifacts from
projects/<old-slug>/…toprojects/<new-slug>/…while preserving history. - Re-keys database rows (
agents,messages,file_reservations) from source to target project. - Records
aliases.jsonunder the target with"former_slugs": [...]for discoverability. - Aborts if agent-name conflicts would break uniqueness in the target (fix names, then retry).
- Idempotent where possible; dry-run always prints a clear plan before apply.
- Requires both projects be in the same repository (validated via
Build slots and helpers (opt-in)
-
amctl envprints helpful environment keys:SLUG,PROJECT_UID,BRANCH,AGENT,CACHE_KEY,ARTIFACT_DIR- Example:
mcp-agent-mail amctl env --path . --agent AliceDev
-
am-runwraps a command with those keys set:- Example:
mcp-agent-mail am-run frontend-build -- npm run dev
- Example:
-
Build slots (advisory, per-project coarse locking):
- Flags:
--ttl-seconds: lease duration (default 3600)--shared/--exclusive: non-exclusive or exclusive lease (default exclusive)--block-on-conflicts: exit non-zero if exclusive conflicts are detected before starting
- Acquire:
- Tool:
acquire_build_slot(project_key, agent_name, slot, ttl_seconds=3600, exclusive=true)
- Tool:
- Renew:
- Tool:
renew_build_slot(project_key, agent_name, slot, extend_seconds=1800)
- Tool:
- Release (non-destructive; marks released):
- Tool:
release_build_slot(project_key, agent_name, slot)
- Tool:
- Notes:
- Slots are recorded under the project archive
build_slots/<slot>/<agent>__<branch>.json exclusive=truereports conflicts if another active exclusive holder exists- Intended for long-running tasks (dev servers, watchers); pair with
am-runandamctl env
- Slots are recorded under the project archive
- Flags:
Product Bus
Group multiple repositories (e.g., frontend, backend, infra) under a single product for product‑wide inbox/search and shared threads.
- Ensure a product:
mcp-agent-mail products ensure MyProduct --name "My Product"
- Link a project (slug or path) into the product:
mcp-agent-mail products link MyProduct .
- Inspect product and linked projects:
mcp-agent-mail products status MyProduct
- Product‑wide message search (FTS):
mcp-agent-mail products search MyProduct "urgent AND deploy" --limit 50
- Product‑wide inbox:
mcp-agent-mail products inbox MyProduct Alice --limit 50 --urgent-only --include-bodies --since-ts "2025-11-01T00:00:00Z"
- Product‑wide thread summarization:
mcp-agent-mail products summarize-thread MyProduct "bd-123" --per-thread-limit 100 --no-llm
Containers
- Build and run locally:
docker build -t mcp-agent-mail . docker run --rm -p 8765:8765 \ -e HTTP_HOST=0.0.0.0 \ -e STORAGE_ROOT=/data/mailbox \ -v agent_mail_data:/data \ mcp-agent-mail - Or with Compose:
docker compose up --build - Notes:
- Runs as an unprivileged user (
appuser, uid 10001). - Includes a HEALTHCHECK against
/health/liveness. - The server reads config from
.envvia python-decouple. You can mount it read-only into the container at/app/.env. - Default bind host is
0.0.0.0in the container; port8765is exposed. - Persistent archive lives under
/data/mailbox(mapped to theagent_mail_datavolume by default).
- Runs as an unprivileged user (
Notes
- A unique
product_uidis stored for each product; you can reference a product by uid or name. - Server tools also exist for orchestration:
ensure_product,products_link,search_messages_product, andresource://product/{key}.
Exclusive file reservations are advisory but visible and auditable:
- A reservation JSON is written to
file_reservations/<sha1(path)>.jsoncapturing holder, pattern, exclusivity, created/expires - The pre-commit guard scans active exclusive reservations and blocks commits that touch conflicting paths held by another agent
- Agents must set
AGENT_NAMEso the guard knows who "owns" the commit - The server continuously evaluates reservations for staleness (agent inactivity + mail/filesystem/git silence) and releases abandoned locks automatically; the
force_release_file_reservationtool uses the same heuristics and notifies the previous holder when another agent clears a stale lease
Install the guard into a code repo (conceptual tool call):
{
"method": "tools/call",
"params": {
"name": "install_precommit_guard",
"arguments": {
"project_key": "/abs/path/backend",
"code_repo_path": "/abs/path/backend"
}
}
}
Configuration
Configuration is loaded from an existing .env via python-decouple. Do not use os.getenv or auto-dotenv loaders.
Changing the HTTP Port
If port 8765 is already in use (e.g., by Cursor's Python extension), you can change it:
Option 1: During installation One-liner with custom port:
curl -fsSL "https://raw.githubusercontent.com/Dicklesworthstone/mcp_agent_mail/main/scripts/install.sh?$(date +%s)" | bash -s -- --port 9000 --yes
Or with local script: ./scripts/install.sh --port 9000 --yes
**Option 2: After installation (CLI)**
```bash
# Change port using CLI command
uv run python -m mcp_agent_mail.cli config set-port 9000
# View current port configuration
uv run python -m mcp_agent_mail.cli config show-port
# Restart server for changes to take effect
scripts/run_server_with_token.sh
Option 3: Manual .env edit
# Edit .env file manually with your text editor (recommended)
nano .env # or vim, code, etc.
# Or append (⚠️ warning: will create duplicate if HTTP_PORT already exists)
echo "HTTP_PORT=9000" >> .env
Option 4: CLI server override
# Override port at server startup (doesn't modify .env)
uv run python -m mcp_agent_mail.cli serve-http --port 9000
from decouple import Config as DecoupleConfig, RepositoryEnv
decouple_config = DecoupleConfig(RepositoryEnv(".env"))
STORAGE_ROOT = decouple_config("STORAGE_ROOT", default="~/.mcp_agent_mail_git_mailbox_repo")
HTTP_HOST = decouple_config("HTTP_HOST", default="127.0.0.1")
HTTP_PORT = int(decouple_config("HTTP_PORT", default=8765))
HTTP_PATH = decouple_config("HTTP_PATH", default="/mcp/")
Common variables you may set:
Configuration reference
| Name | Default | Description |
|---|---|---|
STORAGE_ROOT | ~/.mcp_agent_mail_git_mailbox_repo | Root for per-project repos and SQLite DB |
HTTP_HOST | 127.0.0.1 | Bind host for HTTP transport |
HTTP_PORT | 8765 | Bind port for HTTP transport |
HTTP_PATH | /mcp/ | HTTP path where MCP endpoint is mounted |
HTTP_JWT_ENABLED | false | Enable JWT validation middleware |
HTTP_JWT_SECRET | HMAC secret for HS* algorithms (dev) | |
HTTP_JWT_JWKS_URL | JWKS URL for public key verification | |
HTTP_JWT_ALGORITHMS | HS256 | CSV of allowed algs |
HTTP_JWT_AUDIENCE | Expected aud (optional) | |
HTTP_JWT_ISSUER | Expected iss (optional) | |
HTTP_JWT_ROLE_CLAIM | role | JWT claim name containing role(s) |
HTTP_RBAC_ENABLED | true | Enforce read-only vs tools RBAC |
HTTP_RBAC_READER_ROLES | reader,read,ro | CSV of reader roles |
HTTP_RBAC_WRITER_ROLES | writer,write,tools,rw | CSV of writer roles |
HTTP_RBAC_DEFAULT_ROLE | reader | Role used when none present |
HTTP_RBAC_READONLY_TOOLS | health_check,fetch_inbox,whois,search_messages,summarize_thread | CSV of read-only tool names |
HTTP_RATE_LIMIT_ENABLED | false | Enable token-bucket limiter |
HTTP_RATE_LIMIT_BACKEND | memory | memory or redis |
HTTP_RATE_LIMIT_PER_MINUTE | 60 | Legacy per-IP limit (fallback) |
HTTP_RATE_LIMIT_TOOLS_PER_MINUTE | 60 | Per-minute for tools/call |
HTTP_RATE_LIMIT_TOOLS_BURST | 0 | Optional burst for tools (0=auto=rpm) |
HTTP_RATE_LIMIT_RESOURCES_PER_MINUTE | 120 | Per-minute for resources/read |
HTTP_RATE_LIMIT_RESOURCES_BURST | 0 | Optional burst for resources (0=auto=rpm) |
HTTP_RATE_LIMIT_REDIS_URL | Redis URL for multi-worker limits | |
HTTP_REQUEST_LOG_ENABLED | false | Print request logs (Rich + JSON) |
LOG_JSON_ENABLED | false | Output structlog JSON logs |
INLINE_IMAGE_MAX_BYTES | 65536 | Threshold (bytes) for inlining WebP images during send_message |
CONVERT_IMAGES | true | Convert images to WebP (and optionally inline small ones) |
KEEP_ORIGINAL_IMAGES | false | Also store original image bytes alongside WebP (attachments/originals/) |
LOG_LEVEL | INFO | Server log level |
HTTP_CORS_ENABLED | false | Enable CORS middleware when true |
HTTP_CORS_ORIGINS | CSV of allowed origins (e.g., https://app.example.com,https://ops.example.com) | |
HTTP_CORS_ALLOW_CREDENTIALS | false | Allow credentials on CORS |
HTTP_CORS_ALLOW_METHODS | * | CSV of allowed methods or * |
HTTP_CORS_ALLOW_HEADERS | * | CSV of allowed headers or * |
HTTP_BEARER_TOKEN | Static bearer token (only when JWT disabled) | |
HTTP_ALLOW_LOCALHOST_UNAUTHENTICATED | true | Allow localhost requests without auth (dev convenience) |
HTTP_OTEL_ENABLED | false | Enable OpenTelemetry instrumentation |
OTEL_SERVICE_NAME | mcp-agent-mail | Service name for telemetry |
OTEL_EXPORTER_OTLP_ENDPOINT | OTLP exporter endpoint URL | |
APP_ENVIRONMENT | development | Environment name (development/production) |
DATABASE_URL | sqlite+aiosqlite:///./storage.sqlite3 | SQLAlchemy async database URL |
DATABASE_ECHO | false | Echo SQL statements for debugging |
GIT_AUTHOR_NAME | mcp-agent | Git commit author name |
GIT_AUTHOR_EMAIL | mcp-agent@example.com | Git commit author email |
LLM_ENABLED | true | Enable LiteLLM for thread summaries and discovery |
LLM_DEFAULT_MODEL | gpt-5-mini | Default LiteLLM model identifier |
LLM_TEMPERATURE | 0.2 | LLM temperature for text generation |
LLM_MAX_TOKENS | 512 | Max tokens for LLM completions |
LLM_CACHE_ENABLED | true | Enable LLM response caching |
LLM_CACHE_BACKEND | memory | LLM cache backend (memory or redis) |
LLM_CACHE_REDIS_URL | Redis URL for LLM cache (if backend=redis) | |
LLM_COST_LOGGING_ENABLED | true | Log LLM API costs and token usage |
FILE_RESERVATIONS_CLEANUP_ENABLED | false | Enable background cleanup of expired file reservations |
FILE_RESERVATIONS_CLEANUP_INTERVAL_SECONDS | 60 | Interval for file reservations cleanup task |
FILE_RESERVATION_INACTIVITY_SECONDS | 1800 | Inactivity threshold (seconds) before a reservation is considered stale |
FILE_RESERVATION_ACTIVITY_GRACE_SECONDS | 900 | Grace window for recent mail/filesystem/git activity to keep a reservation active |
FILE_RESERVATIONS_ENFORCEMENT_ENABLED | true | Block message writes on conflicting file reservations |
ACK_TTL_ENABLED | false | Enable overdue ACK scanning (logs/panels; see views/resources) |
ACK_TTL_SECONDS | 1800 | Age threshold (seconds) for overdue ACKs |
ACK_TTL_SCAN_INTERVAL_SECONDS | 60 | Scan interval for overdue ACKs |
ACK_ESCALATION_ENABLED | false | Enable escalation for overdue ACKs |
ACK_ESCALATION_MODE | log | log or file_reservation escalation mode |
ACK_ESCALATION_CLAIM_TTL_SECONDS | 3600 | TTL for escalation file reservations |
ACK_ESCALATION_CLAIM_EXCLUSIVE | false | Make escalation file reservation exclusive |
ACK_ESCALATION_CLAIM_HOLDER_NAME | Ops agent name to own escalation file reservations | |
CONTACT_ENFORCEMENT_ENABLED | true | Enforce contact policy before messaging |
CONTACT_AUTO_TTL_SECONDS | 86400 | TTL for auto-approved contacts (1 day) |
CONTACT_AUTO_RETRY_ENABLED | true | Auto-retry contact requests on policy violations |
MESSAGING_AUTO_REGISTER_RECIPIENTS | true | Automatically create missing local recipients during send_message and retry routing |
MESSAGING_AUTO_HANDSHAKE_ON_BLOCK | true | When contact policy blocks delivery, attempt a contact handshake (auto-accept) and retry |
TOOLS_LOG_ENABLED | true | Log tool invocations for debugging |
LOG_RICH_ENABLED | true | Enable Rich console logging |
LOG_INCLUDE_TRACE | false | Include trace-level logs |
TOOL_METRICS_EMIT_ENABLED | false | Emit periodic tool usage metrics |
TOOL_METRICS_EMIT_INTERVAL_SECONDS | 60 | Interval for metrics emission |
RETENTION_REPORT_ENABLED | false | Enable retention/quota reporting |
RETENTION_REPORT_INTERVAL_SECONDS | 3600 | Interval for retention reports (1 hour) |
RETENTION_MAX_AGE_DAYS | 180 | Max age for retention policy reporting |
QUOTA_ENABLED | false | Enable quota enforcement |
QUOTA_ATTACHMENTS_LIMIT_BYTES | 0 | Max attachment storage per project (0=unlimited) |
QUOTA_INBOX_LIMIT_COUNT | 0 | Max inbox messages per agent (0=unlimited) |
RETENTION_IGNORE_PROJECT_PATTERNS | demo,test*,testproj*,testproject,backendproj*,frontendproj* | CSV of project patterns to ignore in retention/quota reports |
AGENT_NAME_ENFORCEMENT_MODE | coerce | Agent naming policy: strict (reject invalid adjective+noun names), coerce (auto-generate if invalid), always_auto (always auto-generate) |
Development quick start
Prerequisite: complete the setup above (Python 3.14 + uv venv + uv sync).
Dev helpers:
# Quick endpoint smoke test (server must be running locally)
bash scripts/test_endpoints.sh
# Pre-commit guard smoke test (no pytest)
bash scripts/test_guard.sh
Database schema (automatic):
# Tables are created from SQLModel definitions on first run.
# If models change, delete the SQLite DB (and WAL/SHM) and run migrate again.
uv run python -m mcp_agent_mail.cli migrate
Run the server (HTTP-only). Use the Typer CLI or module entry:
uv run python -m mcp_agent_mail.cli serve-http
uv run python -m mcp_agent_mail.http --host 127.0.0.1 --port 8765
Connect with your MCP client using the HTTP (Streamable HTTP) transport on the configured host/port. The endpoint tolerates both /mcp and /mcp/.
Search syntax tips (SQLite FTS5)
- Basic terms:
plan users - Phrase search:
"build plan" - Prefix search:
mig* - Boolean operators:
plan AND users NOT legacy - Field boosting is not enabled by default; subject and body are indexed. Keep queries concise. When FTS is unavailable, the UI/API automatically falls back to SQL LIKE on subject/body.
Design choices and rationale
- HTTP-only FastMCP: Streamable HTTP is the modern remote transport; STDIO is not exposed here by design
- Git + Markdown: Human-auditable, diffable artifacts that fit developer workflows (inbox/outbox mental model)
- SQLite + FTS5: Efficient indexing/search with minimal ops footprint
- Advisory file reservations: Make intent explicit and reviewable; optional guard enforces reservations at commit time
- WebP attachments: Compact images by default; inline embedding keeps small diagrams in context
- Optional: keep original binaries and dedup manifest under
attachments/for audit and reuse
- Optional: keep original binaries and dedup manifest under
Examples (conceptual tool calls)
This section has been removed to keep the README focused. See API Quick Reference below for canonical method signatures.
Operational notes
- One async session per request/task; don't share across concurrent coroutines
- Use explicit loads in async code; avoid implicit lazy loads
- Use async-friendly file operations when needed; Git operations are serialized with a file lock
- Clean shutdown should dispose any async engines/resources (if introduced later)
Security and ops
- Transport
- HTTP-only (Streamable HTTP). Place behind a reverse proxy (e.g., NGINX) with TLS termination for production
- Auth
- Optional JWT (HS*/JWKS) via HTTP middleware; enable with
HTTP_JWT_ENABLED=true - Static bearer token (
HTTP_BEARER_TOKEN) is independent of JWT; when set, BearerAuth protects all routes (including UI). You may use it alone or together with JWT. - When JWKS is configured (
HTTP_JWT_JWKS_URL), incoming JWTs must include a matchingkidheader; tokens withoutkidor with unknownkidare rejected - Starter RBAC (reader vs writer) using role configuration; see
HTTP_RBAC_*settings
- Optional JWT (HS*/JWKS) via HTTP middleware; enable with
- Reverse proxy + TLS (minimal example)
- NGINX location block:
upstream mcp_mail { server 127.0.0.1:8765; } server { listen 443 ssl; server_name mcp.example.com; ssl_certificate /etc/letsencrypt/live/mcp.example.com/fullchain.pem; ssl_certificate_key /etc/letsencrypt/live/mcp.example.com/privkey.pem; location /mcp/ { proxy_pass http://mcp_mail; proxy_set_header Host $host; proxy_set_header X-Forwarded-Proto https; } }
- NGINX location block:
- Backups and retention
- The Git repos and SQLite DB live under
STORAGE_ROOT; back them up together for consistency
- The Git repos and SQLite DB live under
- Observability
- Add logging and metrics at the ASGI layer returned by
mcp.http_app()(Prometheus, OpenTelemetry)
- Add logging and metrics at the ASGI layer returned by
- Concurrency
- Archive writes: per-project
.archive.lockprevents cross-project head-of-line blocking - Commits: repo-level
.commit.lockserializes Git index/commit to avoid races across projects
- Archive writes: per-project
Python client example (HTTP JSON-RPC)
This section has been removed to keep the README focused. Client code samples belong in examples/.
Troubleshooting
- "sender_name not registered"
- Create the agent first with
register_agentorcreate_agent_identity, or check theproject_keyyou're using matches the sender's project
- Create the agent first with
- Pre-commit hook blocks commits
- Set
AGENT_NAMEto your agent identity; release or wait for conflicting exclusive file reservations; inspect.git/hooks/pre-commit
- Set
- Inline images didn't embed
- Ensure
convert_images=true; images are automatically inlined if the compressed WebP size is below the server'sINLINE_IMAGE_MAX_BYTESthreshold (default 64KB). Larger images are stored as attachments instead.
- Ensure
- Message not found
- Confirm the
projectdisambiguation when usingresource://message/{id}; ids are unique per project
- Confirm the
- Inbox empty but messages exist
- Check
since_ts,urgent_only, andlimit; verify recipient names match exactly (case-sensitive)
- Check
FAQ
-
Why Git and SQLite together?
- Git provides human-auditable artifacts and history; SQLite provides fast queries and FTS search. Each is great at what the other isn't.
-
Are file reservations enforced?
- Yes, optionally. The server can block message writes when a conflicting active exclusive reservation exists (
FILE_RESERVATIONS_ENFORCEMENT_ENABLED=true, default). Reservations themselves are advisory and always return bothgrantedandconflicts. The optional pre-commit hook adds local enforcement at commit time in your code repo.
- Yes, optionally. The server can block message writes when a conflicting active exclusive reservation exists (
-
Why HTTP-only?
- Streamable HTTP is the modern remote transport for MCP; avoiding extra transports reduces complexity and encourages a uniform integration path.
-
Why JSON-RPC instead of REST or gRPC?
- MCP defines a tool/resource method call model that maps naturally to JSON-RPC over a single endpoint. It keeps clients simple (one URL, method name + params), plays well with proxies, and avoids SDK lock-in while remaining language-agnostic.
-
Why separate "resources" (reads) from "tools" (mutations)?
- Clear semantics enable aggressive caching and safe prefetch for resources, while tools remain explicit, auditable mutations. This split also powers RBAC (read-only vs writer) without guesswork.
-
Why canonical message storage in Git, not only in the database?
- Git gives durable, diffable, human-reviewable artifacts you can clone, branch, and audit. SQLite provides fast indexing and FTS. The combo preserves governance and operability without a heavyweight message bus.
-
Why advisory file reservations instead of global locks?
- Agents coordinate asynchronously; hard locks create head-of-line blocking and brittle failures. Advisory reservations surface intent and conflicts while the optional pre-commit guard enforces locally where it matters.
-
Why are agent names adjective+noun?
- Memorable identities reduce confusion in inboxes, commit logs, and UI. The scheme yields low collision risk while staying human-friendly (vs GUIDs) and predictable for directory listings.
-
Why is
project_keyan absolute path?- Using the workspace's absolute path creates a stable, collision-resistant project identity across shells and agents. Slugs are derived deterministically from it, avoiding accidental forks of the same project.
-
Why WebP attachments and optional inlining?
- WebP provides compact, high-quality images. Small images can be inlined for readability; larger ones are stored as attachments. You can keep originals when needed (
KEEP_ORIGINAL_IMAGES=true).
- WebP provides compact, high-quality images. Small images can be inlined for readability; larger ones are stored as attachments. You can keep originals when needed (
-
Why both static bearer and JWT/JWKS support?
- Local development should be zero-friction (single bearer). Production benefits from verifiable JWTs with role claims, rotating keys via JWKS, and layered RBAC.
-
Why SQLite FTS5 instead of an external search service?
- FTS5 delivers fast, relevant search with minimal ops. It’s embedded, portable, and easy to back up with the Git archive. If FTS isn’t available, we degrade to SQL LIKE automatically.
-
Why is LLM usage optional?
- Summaries and discovery should enhance, not gate, core functionality. Keeping LLM usage optional controls cost and latency while allowing richer UX when enabled.
API Quick Reference
Tools
Tip: to see tools grouped by workflow with recommended playbooks, fetch
resource://tooling/directory.
| Name | Signature | Returns | Notes |
|---|---|---|---|
health_check | health_check() | {status, environment, http_host, http_port, database_url} | Lightweight readiness probe |
ensure_project | ensure_project(human_key: str) | {id, slug, human_key, created_at} | Idempotently creates/ensures project |
register_agent | register_agent(project_key: str, program: str, model: str, name?: str, task_description?: str, attachments_policy?: str) | Agent profile dict | Creates/updates agent; writes profile to Git |
whois | whois(project_key: str, agent_name: str, include_recent_commits?: bool, commit_limit?: int) | Agent profile dict | Enriched profile for one agent (optionally includes recent commits) |
create_agent_identity | create_agent_identity(project_key: str, program: str, model: str, name_hint?: str, task_description?: str, attachments_policy?: str) | Agent profile dict | Always creates a new unique agent |
send_message | send_message(project_key: str, sender_name: str, to: list[str], subject: str, body_md: str, cc?: list[str], bcc?: list[str], attachment_paths?: list[str], convert_images?: bool, importance?: str, ack_required?: bool, thread_id?: str, auto_contact_if_blocked?: bool) | {deliveries: list, count: int, attachments?} | Writes canonical + inbox/outbox, converts images |
reply_message | reply_message(project_key: str, message_id: int, sender_name: str, body_md: str, to?: list[str], cc?: list[str], bcc?: list[str], subject_prefix?: str) | {thread_id, reply_to, deliveries: list, count: int, attachments?} | Preserves/creates thread, inherits flags |
request_contact | request_contact(project_key: str, from_agent: str, to_agent: str, to_project?: str, reason?: str, ttl_seconds?: int) | Contact link dict | Request permission to message another agent |
respond_contact | respond_contact(project_key: str, to_agent: str, from_agent: str, accept: bool, from_project?: str, ttl_seconds?: int) | Contact link dict | Approve or deny a contact request |
list_contacts | list_contacts(project_key: str, agent_name: str) | list[dict] | List contact links for an agent |
set_contact_policy | set_contact_policy(project_key: str, agent_name: str, policy: str) | Agent dict | Set policy: open, auto, contacts_only, block_all |
fetch_inbox | fetch_inbox(project_key: str, agent_name: str, limit?: int, urgent_only?: bool, include_bodies?: bool, since_ts?: str) | list[dict] | Non-mutating inbox read |
mark_message_read | mark_message_read(project_key: str, agent_name: str, message_id: int) | {message_id, read, read_at} | Per-recipient read receipt |
acknowledge_message | acknowledge_message(project_key: str, agent_name: str, message_id: int) | {message_id, acknowledged, acknowledged_at, read_at} | Sets ack and read |
macro_start_session | macro_start_session(human_key: str, program: str, model: str, task_description?: str, agent_name?: str, file_reservation_paths?: list[str], file_reservation_reason?: str, file_reservation_ttl_seconds?: int, inbox_limit?: int) | {project, agent, file_reservations, inbox} | Orchestrates ensure→register→optional file reservation→inbox fetch |
macro_prepare_thread | macro_prepare_thread(project_key: str, thread_id: str, program: str, model: str, agent_name?: str, task_description?: str, register_if_missing?: bool, include_examples?: bool, inbox_limit?: int, include_inbox_bodies?: bool, llm_mode?: bool, llm_model?: str) | {project, agent, thread, inbox} | Bundles registration, thread summary, and inbox context |
macro_file_reservation_cycle | macro_file_reservation_cycle(project_key: str, agent_name: str, paths: list[str], ttl_seconds?: int, exclusive?: bool, reason?: str, auto_release?: bool) | {file_reservations, released} | File Reservation + optionally release surfaces around a focused edit block |
macro_contact_handshake | `macro_contact_handshake(project_key: str, requester | agent_name: str, target | to_agent: str, to_project?: str, reason?: str, ttl_seconds?: int, auto_accept?: bool, welcome_subject?: str, welcome_body?: str)` |
search_messages | search_messages(project_key: str, query: str, limit?: int) | list[dict] | FTS5 search (bm25) |
summarize_thread | summarize_thread(project_key: str, thread_id: str, include_examples?: bool, llm_mode?: bool, llm_model?: str, per_thread_limit?: int) | Single: {thread_id, summary, examples} Multi (comma-sep): {threads[], aggregate} | Extracts participants, key points, actions. Use comma-separated thread_id for multi-thread digest. |
install_precommit_guard | install_precommit_guard(project_key: str, code_repo_path: str) | {hook} | Install a Git pre-commit guard in a target repo |
uninstall_precommit_guard | uninstall_precommit_guard(code_repo_path: str) | {removed} | Remove the guard from a repo |
file_reservation_paths | file_reservation_paths(project_key: str, agent_name: str, paths: list[str], ttl_seconds?: int, exclusive?: bool, reason?: str) | {granted: list, conflicts: list} | Advisory leases; Git artifact per path |
release_file_reservations | release_file_reservations(project_key: str, agent_name: str, paths?: list[str], file_reservation_ids?: list[int]) | {released, released_at} | Releases agent's active file reservations |
force_release_file_reservation | force_release_file_reservation(project_key: str, agent_name: str, file_reservation_id: int, notify_previous?: bool, note?: str) | {released, released_at, reservation} | Clears stale reservations using inactivity/mail/fs/git heuristics and notifies the previous holder |
renew_file_reservations | renew_file_reservations(project_key: str, agent_name: str, extend_seconds?: int, paths?: list[str], file_reservation_ids?: list[int]) | {renewed, file reservations[]} | Extend TTL of existing file reservations |
Resources
| URI | Params | Returns | Notes |
|---|---|---|---|
resource://config/environment | — | {environment, database_url, http} | Inspect server settings |
resource://tooling/directory | — | {generated_at, metrics_uri, clusters[], playbooks[]} | Grouped tool directory + workflow playbooks |
resource://tooling/schemas | — | {tools: {<name>: {required[], optional[], aliases{}}}} | Argument hints for tools |
resource://tooling/metrics | — | {generated_at, tools[]} | Aggregated call/error counts per tool |
resource://tooling/locks | — | {locks[], summary} | Active locks and owners (debug only). Categories: archive (per-project .archive.lock) and custom (e.g., repo .commit.lock). |
resource://tooling/capabilities/{agent}{?project} | listed | {generated_at, agent, project, capabilities[]} | Capabilities assigned to the agent (see deploy/capabilities/agent_capabilities.json) |
resource://tooling/recent/{window_seconds}{?agent,project} | listed | {generated_at, window_seconds, count, entries[]} | Recent tool usage filtered by agent/project |
resource://projects | — | list[project] | All projects |
resource://project/{slug} | slug | {project..., agents[]} | Project detail + agents |
resource://file_reservations/{slug}{?active_only} | slug, active_only? | list[file reservation] | File reservations plus staleness metadata (heuristics, last activity timestamps) |
resource://message/{id}{?project} | id, project | message | Single message with body |
resource://thread/{thread_id}{?project,include_bodies} | thread_id, project, include_bodies? | {project, thread_id, messages[]} | Thread listing |
resource://inbox/{agent}{?project,since_ts,urgent_only,include_bodies,limit} | listed | {project, agent, count, messages[]} | Inbox listing |
resource://mailbox/{agent}{?project,limit} | project, limit | {project, agent, count, messages[]} | Mailbox listing (recent messages with basic commit ref) |
resource://mailbox-with-commits/{agent}{?project,limit} | project, limit | {project, agent, count, messages[]} | Mailbox listing enriched with commit metadata |
resource://outbox/{agent}{?project,limit,include_bodies,since_ts} | listed | {project, agent, count, messages[]} | Messages sent by the agent |
resource://views/acks-stale/{agent}{?project,ttl_seconds,limit} | listed | {project, agent, ttl_seconds, count, messages[]} | Ack-required older than TTL without ack |
resource://views/urgent-unread/{agent}{?project,limit} | listed | {project, agent, count, messages[]} | High/urgent importance messages not yet read |
resource://views/ack-required/{agent}{?project,limit} | listed | {project, agent, count, messages[]} | Pending acknowledgements for an agent |
resource://views/ack-overdue/{agent}{?project,ttl_minutes,limit} | listed | {project, agent, ttl_minutes, count, messages[]} | Ack-required older than TTL without ack |
Client Integration Guide
- Fetch onboarding metadata first. Issue
resources/readforresource://tooling/directory(and optionallyresource://tooling/metrics) before exposing tools to an agent. Use the returned clusters and playbooks to render a narrow tool palette for the current workflow rather than dumping every verb into the UI. - Scope tools per workflow. When the agent enters a new phase (e.g., "Messaging Lifecycle"), remount only the cluster's tools in your MCP client. This mirrors the workflow macros already provided and prevents "tool overload."
- Monitor real usage. Periodically pull or subscribe to log streams containing the
tool_metrics_snapshotevents emitted by the server (or queryresource://tooling/metrics) so you can detect high-error-rate tools and decide whether to expose macros or extra guidance. - Fallback to macros for smaller models. If you're routing work to a lightweight model, prefer the macro helpers (
macro_start_session,macro_prepare_thread,macro_file_reservation_cycle,macro_contact_handshake) and hide the granular verbs until the agent explicitly asks for them. - Show recent actions. Read
resource://tooling/recent/60?agent=<name>&project=<slug>(adjust window as needed) to display the last few successful tool invocations relevant to the agent/project.
See examples/client_bootstrap.py for a runnable reference implementation that applies the guidance above.
{
"steps": [
"resources/read -> resource://tooling/directory",
"select active cluster (e.g. messaging)",
"mount tools listed in cluster.tools plus macros if model size <= S",
"optional: resources/read -> resource://tooling/metrics for dashboard display",
"optional: resources/read -> resource://tooling/recent/60?agent=<name>&project=<slug> for UI hints"
]
}
Monitoring & Alerts
- Enable metric emission. Set
TOOL_METRICS_EMIT_ENABLED=trueand choose an interval (TOOL_METRICS_EMIT_INTERVAL_SECONDS=120is a good starting point). The server will periodically emit a structured log entry such as:
{
"event": "tool_metrics_snapshot",
"tools": [
{"name": "send_message", "cluster": "messaging", "calls": 42, "errors": 1},
{"name": "file_reservation_paths", "cluster": "file reservations", "calls": 11, "errors": 0}
]
}
- Ship the logs. Forward the structured stream (stderr/stdout or JSON log files) into your observability stack (e.g., Loki, Datadog, Elastic) and parse the
tools[]array. - Alert on anomalies. Create a rule that raises when
errors / callsexceeds a threshold for any tool (for example 5% over a 5-minute window) so you can decide whether to expose a macro or improve documentation. - Dashboard the clusters. Group by
clusterto see where agents are spending time and which workflows might warrant additional macros or guard-rails.
See docs/observability.md for a step-by-step cookbook (Loki/Prometheus example pipelines included), and docs/GUIDE_TO_OPTIMAL_MCP_SERVER_DESIGN.md for a comprehensive design guide covering tool curation, capability gating, security, and observability best practices.
Operations teams can follow docs/operations_alignment_checklist.md, which links to the capability templates in deploy/capabilities/ and the sample Prometheus alert rules in deploy/observability/.
Deployment quick notes
- Direct uvicorn:
uvicorn mcp_agent_mail.http:build_http_app --factory --host 0.0.0.0 --port 8765 - Python module:
python -m mcp_agent_mail.http --host 0.0.0.0 --port 8765 - Gunicorn:
gunicorn -c deploy/gunicorn.conf.py mcp_agent_mail.http:build_http_app --factory - Docker:
docker compose up --build
CI/CD
- Lint and Typecheck CI: GitHub Actions workflow runs Ruff and Ty on pushes/PRs to main/develop.
- Release: Pushing a tag like
v0.1.0builds and pushes a multi-arch Docker image to GHCR underghcr.io/<owner>/<repo>withlatestand version tags. - Nightly: A scheduled workflow runs migrations and lists projects daily for lightweight maintenance visibility.
Log rotation (optional)
If not using journald, a sample logrotate config is provided at deploy/logrotate/mcp-agent-mail to rotate /var/log/mcp-agent-mail/*.log weekly, keeping 7 rotations.
Logging (journald vs file)
- Default systemd unit (
deploy/systemd/mcp-agent-mail.service) is configured to send logs to journald (StandardOutput/StandardError=journal). - For file logging, configure your process manager to write to files under
/var/log/mcp-agent-mail/*.logand install the provided logrotate config. - Environment file path for systemd is
/etc/mcp-agent-mail.env(seedeploy/systemd/mcp-agent-mail.service).
Container build and multi-arch push
Use Docker Buildx for multi-arch images. Example flow:
# Create and select a builder (once)
docker buildx create --use --name mcp-builder || docker buildx use mcp-builder
# Build and test locally (linux/amd64)
docker buildx build --load -t your-registry/mcp-agent-mail:dev .
# Multi-arch build and push (amd64, arm64)
docker buildx build \
--platform linux/amd64,linux/arm64 \
-t your-registry/mcp-agent-mail:latest \
-t your-registry/mcp-agent-mail:v0.1.0 \
--push .
Recommended tags: a moving latest and immutable version tags per release. Ensure your registry login is configured (docker login).
Systemd manual deployment steps
- Copy project files to
/opt/mcp-agent-mailand ensure permissions (ownerappuser). - Place environment file at
/etc/mcp-agent-mail.envbased ondeploy/env/production.env. - Install service file
deploy/systemd/mcp-agent-mail.serviceto/etc/systemd/system/. - Reload systemd and start:
sudo systemctl daemon-reload
sudo systemctl enable mcp-agent-mail
sudo systemctl start mcp-agent-mail
sudo systemctl status mcp-agent-mail
Optional (non-journald log rotation): install deploy/logrotate/mcp-agent-mail into /etc/logrotate.d/ and write logs to /var/log/mcp-agent-mail/*.log via your process manager or app config.
See deploy/gunicorn.conf.py for a starter configuration. For project direction and planned areas, read project_idea_and_guide.md.
CLI Commands
The project exposes a developer CLI for common operations:
serve-http: run the HTTP transport (Streamable HTTP only)migrate: ensure schema and FTS structures existlint/typecheck: developer helperslist-projects [--include-agents]: enumerate projectsguard install <project_key> <code_repo_path>: install the pre-commit guard into a repoguard uninstall <code_repo_path>: remove the guard from a reposhare wizard: launch interactive deployment wizard (auto-installs CLIs, authenticates, exports, deploys to GitHub Pages or Cloudflare Pages)share export --output <path> [options]: export mailbox to a static HTML bundle (see Static Mailbox Export section for full options)share update <bundle_path> [options]: refresh an existing bundle using recorded (or overridden) export settingsshare preview <bundle_path> [--port N] [--open-browser]: serve a static bundle locally for inspectionshare verify <bundle_path> [--public-key <key>]: verify bundle integrity (SRI hashes and Ed25519 signature)share decrypt <encrypted_path> [--identity <file> | --passphrase]: decrypt an age-encrypted bundleconfig set-port <port>: change the HTTP server port (updates .env)config show-port: display the current configured HTTP portclear-and-reset-everything [--force] [--archive/--no-archive]: DELETE the SQLite database (incl. WAL/SHM) and WIPE all contents underSTORAGE_ROOTafter optionally saving a restore point. Without flags it prompts to create an archive first;--force --no-archiveskips all prompts for automation.list-acks --project <key> --agent <name> [--limit N]: list messages requiring acknowledgement for an agent where ack is missingacks pending <project> <agent> [--limit N]: show pending acknowledgements for an agentacks remind <project> <agent> [--min-age-minutes N] [--limit N]: highlight pending ACKs older than a thresholdacks overdue <project> <agent> [--ttl-minutes N] [--limit N]: list overdue ACKs beyond TTLfile_reservations list <project> [--active-only/--no-active-only]: list file reservationsfile_reservations active <project> [--limit N]: list active file reservationsfile_reservations soon <project> [--minutes N]: show file reservations expiring soondoctor check [PROJECT] [--verbose] [--json]: run comprehensive diagnostics on mailbox healthdoctor repair [PROJECT] [--dry-run] [--yes] [--backup-dir PATH]: semi-automatic repair with backup before changesdoctor backups [--json]: list available diagnostic backupsdoctor restore <backup_path> [--dry-run] [--yes]: restore from a diagnostic backup
Examples:
# Interactive wizard: export + deploy to GitHub Pages (easiest)
./scripts/share_to_github_pages.py
# Export a static bundle with signing and encryption
uv run python -m mcp_agent_mail.cli share export \
--output ./bundle \
--signing-key ./keys/signing.key \
--age-recipient age1abc...xyz
# Preview a bundle locally
uv run python -m mcp_agent_mail.cli share preview ./bundle --port 9000 --open-browser
# Verify bundle integrity
uv run python -m mcp_agent_mail.cli share verify ./bundle
# Refresh an existing bundle in place with recorded settings
uv run python -m mcp_agent_mail.cli share update ./bundle
# Change server port
uv run python -m mcp_agent_mail.cli config set-port 9000
# Install guard into a repo
uv run python -m mcp_agent_mail.cli guard install /abs/path/backend /abs/path/backend
# List pending acknowledgements for an agent
uv run python -m mcp_agent_mail.cli acks pending /abs/path/backend BlueLake --limit 10
# Run mailbox health diagnostics
uv run python -m mcp_agent_mail.cli doctor check
# Preview repairs without making changes
uv run python -m mcp_agent_mail.cli doctor repair --dry-run
# Run repairs (creates backup first, prompts for data changes)
uv run python -m mcp_agent_mail.cli doctor repair
# List available backups
uv run python -m mcp_agent_mail.cli doctor backups
# Restore from a backup
uv run python -m mcp_agent_mail.cli doctor restore /path/to/backup --dry-run
# WARNING: Destructive reset (clean slate)
uv run python -m mcp_agent_mail.cli clear-and-reset-everything --force
Client integrations
Use the automated installer to wire up supported tools automatically (e.g., Claude Code, Cline, Windsurf, OpenCode). Run scripts/automatically_detect_all_installed_coding_agents_and_install_mcp_agent_mail_in_all.sh or the one-liner in the Quickstart above.
Tool-specific integration scripts
For manual integration or customization, dedicated scripts are available:
| Tool | Script | What it configures |
|---|---|---|
| Claude Code | scripts/integrate_claude_code.sh | .claude/settings.json, hooks, MCP server |
| Codex CLI | scripts/integrate_codex_cli.sh | ~/.codex/config.toml, MCP server, notify handler |
| Gemini CLI | scripts/integrate_gemini_cli.sh | ~/.gemini/settings.json, MCP server, hooks |
Each script:
- Detects the MCP server endpoint from your settings
- Generates or reuses a bearer token for authentication
- Configures the MCP server connection
- Installs hooks/notify handlers for inbox reminders
- Bootstraps your project and agent identity on the server
Automatic inbox reminders
Agents often get absorbed in their work and forget to check their mail. The integration scripts install lightweight hooks that periodically remind agents when they have unread messages.
How it works:
- A rate-limited hook script (
scripts/hooks/check_inbox.sh) runs after certain tool invocations - It checks the inbox via a fast curl call (avoids Python import overhead)
- If there are unread messages, it outputs a brief reminder
- Rate limited to at most once per 2 minutes to avoid noise
Claude Code / Gemini CLI:
The hook is configured as a PostToolUse hook that fires after Bash or shell tool invocations:
{
"hooks": {
"PostToolUse": [
{
"matcher": "Bash",
"hooks": [{ "type": "command", "command": "...check_inbox.sh" }]
}
]
}
}
Codex CLI:
Uses the top-level notify configuration in config.toml (must appear before any [section] headers) to fire on agent-turn-complete events:
notify = ["/path/to/.codex/hooks/notify_wrapper.sh"]
Additional hooks (Claude Code only):
| Event | What it does |
|---|---|
SessionStart | Shows active file reservations and pending acknowledgments |
PreToolUse (Edit) | Warns about file reservations expiring within 10 minutes |
PostToolUse (send_message) | Lists recent acknowledgment requests |
PostToolUse (file_reservation_paths) | Shows current file reservations |
Environment variables for hooks
The inbox check hooks accept these environment variables (set automatically by the integration scripts):
| Variable | Description | Default |
|---|---|---|
AGENT_MAIL_PROJECT | Project key (absolute path) | required |
AGENT_MAIL_AGENT | Agent name | required |
AGENT_MAIL_URL | Server URL | http://127.0.0.1:8765/mcp/ |
AGENT_MAIL_TOKEN | Bearer token | none |
AGENT_MAIL_INTERVAL | Seconds between checks | 120 |
About Contributions: Please don't take this the wrong way, but I do not accept outside contributions for any of my projects. I simply don't have the mental bandwidth to review anything, and it's my name on the thing, so I'm responsible for any problems it causes; thus, the risk-reward is highly asymmetric from my perspective. I'd also have to worry about other "stakeholders," which seems unwise for tools I mostly make for myself for free. Feel free to submit issues, and even PRs if you want to illustrate a proposed fix, but know I won't merge them directly. Instead, I'll have Claude or Codex review submissions via
ghand independently decide whether and how to address them. Bug reports in particular are welcome. Sorry if this offends, but I want to avoid wasted time and hurt feelings. I understand this isn't in sync with the prevailing open-source ethos that seeks community contributions, but it's the only way I can move at this velocity and keep my sanity.