agentleFS
ryanmacleanryanmaclean/skills/bop-catalog/SKILL.md

bop-catalog

Skill1 starsChanged 49 days agoMITView on GitHub

Discover callable bop cards via the fleet catalog endpoint. Any LM can list available capabilities, read their schemas, and invoke them by id. Trigger phrases — bop catalog, discover card, list cards, available capabilities, fleet capabilities.

The file

Files
SKILL.md

Bop Catalog — AX Discovery Endpoint

Phase 18 Wave 1 artifact. Turns bop from tribal to self-describing.

Single stable endpoint

http://10.0.3.210:3000/string/bop/releases/download/catalog%2Flatest/latest.json

The catalog/latest release tag is a rolling pointer. Versioned snapshots live at catalog/v1-YYYYMMDD tags on the same repo (string/bop on QNAS Gitea).

Usage (<=3 tool calls)

  1. List — what capabilities exist?
    curl -s <endpoint> | jq '.cards[] | {id, title, min_trust, side_effects}'
  2. Inspect — full manifest for one card (embedded in the catalog, no extra fetch needed):
    curl -s <endpoint> | jq '.cards[] | select(.id == "example-hello") | .manifest'
  3. Invoke — dispatch with validated input:
    bop submit example-hello --input '{"name": "agent"}'

Schemas

  • ~/bop/schema/manifest.v1.json — per-card manifest (id, version, inputs, outputs, min_trust, side_effects, model_hints, examples)
  • ~/bop/schema/catalog.v1.json — aggregate catalog (schema_version, generated_at, cards[])

Both validate as JSON Schema draft 2020-12.

Regenerating the catalog

nu ~/bop/scripts/generate-catalog.nu
tea release create --login qnas --repo string/bop \
  --tag "catalog/v1-$(date +%Y%m%d)" \
  --asset ~/bop/catalog/latest.json \
  --note "..."

Then re-upload latest.json to the catalog/latest rolling tag.

Versioning

  • v1 is the current schema. Changes that break LM parsing bump to v2 with a new endpoint path. v1 stays working.
  • Per-card version follows SemVer.
  • catalog/latest tag is a rolling pointer; dated catalog/v1-YYYYMMDD tags give point-in-time snapshots.

Trust tiers

Cards declare min_trust: L0|L1|L2|L3. The dispatcher must refuse to invoke a card for an agent below that tier. See fleet trust-tiers docs.

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.