agentleFS
Sign inStart free
pankaj-mahaurpankaj-mahaur/Agent-Daemon/AGENTS.md

Agent-Daemon

AGENTS.md0 starsChanged 7 days agoMITView on GitHub

This project uses agent-daemon for multi-agent orchestration and self-improving memory. Memory files with {{PLACEHOLDER}} text haven't been bootstrapped yet. The digest pipeline keeps memory updated automatically after bootstrapping. Use multi-agent when ALL of these are true: - Task spans 2+ domains (backend + frontend, code + tests, etc.) - Subtasks can run in parallel - Estimated work exceeds ~30 minutes Stay single-agent for: quick fixes, single-file changes, questions, reviews of small scope. Keep the docs, CHANGELOG and skills/daemon/* in step…

The file

We split this AGENTS.md into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use multi-agent when ALL of these are true:

    line 38
  2. Very specific:

    Keep the docs, CHANGELOG and skills/daemon/* in step with the code, and push nothing personal or internal (see .gitignore).

    line 64

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

Agent Daemon

This project uses agent-daemon for multi-agent orchestration and self-improving memory.

Getting Started

  1. Scaffold: ad init — creates .agent-daemon/memory/ templates, AD-INSTRUCTIONS.md (the operating manual) and a managed section in CLAUDE.md
  2. Bootstrap: Tell Claude "bootstrap the daemon memory" — reads your project and populates memory with real context
  3. Verify: ad doctor — confirms everything is wired up

Memory files with {{PLACEHOLDER}} text haven't been bootstrapped yet. The digest pipeline keeps memory updated automatically after bootstrapping.

Quick Reference

CommandAliasWhat it does
ad doctorVerify install — hooks, PATH, settings, harness login and sandbox
ad initScaffold .agent-daemon/ in this project
adOpen the terminal UI (ad's own agent harness; signs in on first run)
ad team list-templatesad ttShow available team templates
ad team create --template <name> --task "..."ad tcCreate a team with roles + task graph
ad team status --team <id>ad tsKanban board — tasks, agents, progress
ad team inbox --team <id> --agent <name>ad tiRead completion messages
ad spawn --team <id> --role <role> --task "..." --cwd .ad spLaunch a worker in an isolated worktree (a sandboxed worker on ad's engine by default; --engine claude for headless claude)
ad team delete --team <id>ad tdRemove a team
ad team cleanupad tuPrune stale worktrees

Templates

  • solo-with-qa — 1 dev + 1 QA (simplest, good for most tasks)
  • full-stack-feature — lead + backend + frontend + QA (parallel work)
  • bug-triage-team — lead + investigator + fixer + reviewer
  • code-review-team — lead + security + performance reviewers

When to use multi-agent

Use multi-agent when ALL of these are true:

  • Task spans 2+ domains (backend + frontend, code + tests, etc.)
  • Subtasks can run in parallel
  • Estimated work exceeds ~30 minutes

Stay single-agent for: quick fixes, single-file changes, questions, reviews of small scope.

Workflow

  1. Analyze — Break the task into subtasks with dependencies
  2. Ask the user — Present the team plan (template, roles, task graph) and wait for approval
  3. Create — ad tc --template <name> --task "..."
  4. Spawn — ad sp --team <id> --role <role> --task "..." --cwd . for each worker
  5. Monitor — ad ts --team <id> to check progress, ad ti for inbox
  6. Merge — Review agent branches, merge in dependency order, test
  7. Cleanup — ad td --team <id> when done

Rules

  • Always ask before spawning — Never create teams or spawn agents without showing the plan and getting user approval first.
  • Leader = your session — You (the agent in this session) are the team lead. Workers run in background worktrees.
  • Isolation — Each agent works on its own git branch. Changes don't touch the user's working branch until explicitly merged.
  • Review before merge — Always let the user review agent output before merging branches.

Before every push

Keep the docs, CHANGELOG and skills/daemon/* in step with the code, and push nothing personal or internal (see .gitignore). The full checklist is .claude/skills/ad-pre-push/SKILL.md; verify with node runtime/scripts/check-doc-links.mjs and cd runtime && npm test.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free