agentleFS
hushh-labshushh-labs/hushh-research/AGENTS.md

hushh-research

AGENTS.md40 starsChanged 3 days agoApache-2.0View on GitHub

Precision. Discretion. Duty of care. “An agent should work for the person whose life it touches.” Build One around that personal-agent identity: owner loyalty, explicit consent, verified outcomes and thoughtful restraint. Read .agents/instructions.md before routing, planning, delegating, or changing this repository. Its complete operating contract is binding for this repository and inherited by every development subagent. Package-local instructions refine that contract within their scope. Every coding agent uses this same entrypoint and reads the canonical sources directly. Keep one concise…

The file

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

8 rules. About 0 are common, 4 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read .agents/instructions.md before routing, planning, delegating, or changing this repository.

    line 9
  2. Distinctive:

    Keep one concise current source for documentation and agent guidance.

    line 16
  3. Very specific:

    All development sources belong in .agents/:

    line 25
  4. Very specific:

    Create or evolve these through the existing governance workflow.

    line 33
  5. Distinctive:

    Run independent work in bounded parallel lanes under the canonical delegation contract.

    line 40
  6. Distinctive:

    After handoff, verify owned-process and temporary-resource cleanup and close child sessions when supported.

    line 42
  7. Very specific:

    Start with the development-system map, then the existing ./bin/hushh agents route-task <workflow-id> command for the owning workflow.

    line 48
  8. Distinctive:

    Before changing a readiness check, probe, grader or instruction text that a live check grades, follow the readiness guardrails.

    line 54

≈ “Common” is an estimate: at least 1 in 5 of the 41,581 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

Hussh · Agent Engineering

Precision. Discretion. Duty of care.

“An agent should work for the person whose life it touches.” Build One around that personal-agent identity: owner loyalty, explicit consent, verified outcomes and thoughtful restraint.

Read .agents/instructions.md before routing, planning, delegating, or changing this repository. Its complete operating contract is binding for this repository and inherited by every development subagent.

Package-local instructions refine that contract within their scope. Every coding agent uses this same entrypoint and reads the canonical sources directly.

Keep one concise current source for documentation and agent guidance. Do not create or retain stale, duplicate, or bloated copies. Preserve unique uncommitted work and required evidence before removing a copy; use Git history for prior versions instead of checking in historical snapshots.

Generated review/audit evidence belongs in ignored local or CI artifacts, not source commits. Preserve runtime contracts and required app assets under the canonical evidence policy.

All development sources belong in .agents/:

  • skills: reusable procedures and their supporting tools.
  • workflows: ownership, routing and verification.
  • subagents: the unified read-only engineering fleet.
  • evals: routing and engineering evaluations.
  • configuration: bounded execution and tool connections.

Create or evolve these through the existing governance workflow. Do not create .claude, .codex, .qwen or another vendor-specific instruction, skill or agent tree. Do not generate per-host adapters or copies of the development fleet. Root AGENTS.md and .agents are the portable operating contract. Tool access and automatic discovery still depend on the active host; never claim that reading this file registers MCP servers or subagents automatically.

Run independent work in bounded parallel lanes under the canonical delegation contract. After handoff, verify owned-process and temporary-resource cleanup and close child sessions when supported. Permanently remove task-created temporary worktrees, including worktrees branched from a base branch, after successful verification and handoff under the local resource and process lifecycle gate; completion alone is not proof of cleanup.

Start with the development-system map, then the existing ./bin/hushh agents route-task <workflow-id> command for the owning workflow. Route CI work through the existing ci-watch-and-heal or release-readiness workflow and the repo-operations skill; keep required checks and release authority intact.

Before changing a readiness check, probe, grader or instruction text that a live check grades, follow the readiness guardrails.

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.