@diegosouzapw/swarm-mail
BCoordinate with other agents using Swarm mail and file locking
Install
agr install @diegosouzapw/swarm-mail --target claudeWrites 2 files into .claude/skills/, pinned to git-7167949f.
- .claude/skills/swarm-mail/SKILL.md
- .claude/skills/swarm-mail/metadata.json
Document
name: swarm-mail description: Coordinate with other agents using Swarm mail and file locking globs:
- "**/*" alwaysApply: true
Swarm Multi-Agent Coordination
You are part of a Swarm deployment with multiple AI agents working on the same codebase. Use these patterns to coordinate effectively.
When to Use Mail vs Queue vs Direct
- Mail: Handoff tasks to specific agents, request reviews, report completion
- Queue: Send work to yourself for later (cooldown, complex sequences)
- Direct: Only for emergency interrupts (use
swarm inject)
Writing Actionable Handoff Messages
Good handoff messages require no follow-up questions.
Good Example
Subject: Review PR #123 - User authentication refactor Body:
- PR is ready for review at https://github.com/...
- Focus on: error handling in login flow
- Tests pass locally, CI pending
- After review, send results to agent-xyz
Bad Example
Subject: Please review Body: The PR is ready.
Advisory File Locking
Before editing files, claim a lock:
swarm lock claim --agent $AGENT_ID --path "src/api/auth.go" --ttl 30m
Check for conflicts before claiming:
swarm lock check --path "src/api/auth.go"
Always release when done:
swarm lock release --agent $AGENT_ID
Subject/Body Conventions
Subjects should be:
- Specific: "Fix null pointer in UserService.getById()"
- Searchable: Include file names, function names, issue numbers
- Actionable: Start with verb (Fix, Review, Implement, Update)
Checking Your Inbox
Poll for new work regularly:
swarm mail inbox --agent $AGENT_ID --unread
Additional template files
See the templates in templates/:
- templates/handoff.md
- templates/review-request.md
- templates/conflict-resolution.md
Trustgrade B
- passBody integrity
Whether the stored document is plausibly the kind of file the artifact declares, rather than something fetched by mistake.
- passType matchnot applicable to this artifact type
Whether the artifact is really the kind of thing its metadata claims it is.
- passFreshness
How long since the source repository was last pushed to.
- passPrompt injection
Scans the artifact's own text for instructions aimed at your agent rather than at you.
- warnLicenseno SPDX license detected
Whether the source repository declares an SPDX license permissive enough to redistribute.
How the grade is calculated
Each check contributes 0 points when it passes, 1 when it warns, and 2 when it fails. The total maps to a letter:
- Aevery check passed
- Bone warning
- Ctwo warnings
- Dprompt injection or body integrity failed, or three warnings
- Fone of those failed, and something else is wrong
These are automated hygiene checks, not a security audit, and not a dependency or vulnerability scan. A grade of A means nothing was flagged — not that the artifact is safe.
Versions
git-7167949f27e72026-07-31