agentleFS
Sign inStart free
PuntoyComaTechPuntoyComaTech/hermes-agent-squads/AGENTS.md

hermes-agent-squads

AGENTS.md3 starsChanged 18 days agoMITView on GitHub

Contributor contract for this repository. Humans start at README.md; Hermes behavior is documented in skills/hermes-platform/references/. The builder's profile distribution for Hermes Agent squads: manifests (squads/*/squad.yaml, builder.yaml), skills, tested scripts, a schema. Installed with hermes profile install, reconciled by scripts/squadctl.py. A user asks for this with "Install github.com/PuntoyComaTech/hermes-agent-squads following its AGENTS.md". The agent runs only hermes profile install github.com/PuntoyComaTech/hermes-agent-squads --name builder --alias, then tells the user, in this order: 1. hermes -p builder model: pick the builder's model. 2. builder chat…

  • Reads credentials

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:

    Do not reimplement them.

    line 30
  2. Very specific:

    No secrets: never read or write .env, auth.json or credential values.

    line 40

≈ “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

AGENTS.md

Contributor contract for this repository. Humans start at README.md; Hermes behavior is documented in skills/hermes-platform/references/.

What this is

The builder's profile distribution for Hermes Agent squads: manifests (squads/*/squad.yaml, builder.yaml), skills, tested scripts, a schema. Installed with hermes profile install, reconciled by scripts/squadctl.py.

Installing for a user

A user asks for this with "Install github.com/PuntoyComaTech/hermes-agent-squads following its AGENTS.md". The agent runs only hermes profile install github.com/PuntoyComaTech/hermes-agent-squads --name builder --alias, then tells the user, in this order:

  1. hermes -p builder model: pick the builder's model.
  2. builder chat (or hermes -p builder chat) and say "set me up"; the builder asks which squads to install and which model each bot uses.
  3. Desktop app: Sessions, pick a squad manager in the bottom selector, "+" creates a project.

The agent never installs squads or picks models for the user. Updates then run daily on their own (cron builder-update, scripts/auto_update.py); "update yourself" to the builder runs the same update now.

Rules

  • Use the tools through scripts/hpy <tool>.py (the Hermes runtime interpreter): squadctl.py applies changes, skill_lint.py gates skills, model_advisor.py assigns models, kanban_chain.py builds chains, project_sync.py copies a manager's projects to its workers, auto_update.py updates the builder and squads. Do not reimplement them.
  • English only in repo content; bots speak the user's language at runtime.
  • Model ids only in scripts/data/model-roles.yaml and local/installed.yaml (enforced).
  • Real scripts, never prose scripts: Python for YAML/JSON/HTTP work, --help on every script; bash only for the shims and one-line cron wrappers.
  • Skills pass scripts/hpy skill_lint.py; rules: skill_lint.py --rules.
  • Verify every platform claim against local Hermes docs or source before writing it; when Hermes contradicts the repo, Hermes wins: fix the repo. Bot-loaded files (SOUL.md, skills/, squads/) state facts without source paths, decision ids or history.
  • Structural decisions: CONTRIBUTING.md -> Decisions.
  • No secrets: never read or write .env, auth.json or credential values.
  • Plan, user OK, apply for anything that touches a profile; profile export before any prune or uninstall.
  • Community plugins and hub skills: skills/finding-capabilities steps 3-5. Squad rules (naming, visibility, topology, credentials): skills/designing-squads -> Rules, Topology.

Map

PathWhat
distribution.yaml, config.yaml, SOUL.md, builder.yamlshipped profile payload and builder bot spec
skills/builder skills; hermes-platform is the verified reference
scripts/, scripts/data/squadctl and helpers; lane tables
squads/<key>/one squad: squad.yaml, souls/, skills/, scripts/, README
local/ (installed builder only, never shipped)installed.yaml, overlays and local squads in squads/<key>/, backups, state
squads/_schema/squad.schema.jsonmanifest schema
squads/_plugins/plugins squadctl installs: shared-browser-tabs (agent half), squad-hub (agent half on every manager, Desktop half opt-in)
docs/decisions.mdcondensed decision log
tests/pytest against a temporary HERMES_HOME

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