agentleFS
TheVellerTheVeller/claude-skills/skills/agent-operations/drift-hunt-roster/SKILL.md

drift-hunt-roster

Skill4 starsChanged 14 days agoMITView on GitHub

Use when running a daily roster drift hunt - compare live agents and rooms on the box against the fleet inventory docs and report only new deltas to an ops room.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If there is a delta, write one finding signed with the bot's name: 🎯 what changed · 🔗 one link or n/d · ⭐ score · ➡️ next owner and …

    line 11
  2. Distinctive:

    If there is no delta, stay silent.

    line 12
  3. Distinctive:

    If a run fails or skips, say so in the next stand-up and re-verify before reposting.

    line 14

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill 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
SKILL.md

Roster drift hunt

  1. Count live dirs in the agents folder on the box. A dir with group.json is a room; any other dir is an agent.
  2. Compare against the source-of-truth docs (inventory and roster) and the locked expected count kept in the routine prompt.
  3. For each id not in the docs or the closed list, read its profile (name, description, manager) and its creation time with stat -c %w (birth time). Never use mtime for creation dates; bulk mtime changes from syncs or restores are not drift.
  4. A delta is only: a new live id not listed, a listed id missing from the box, a new rename, or the canon count rewritten without a lock. Items already reported and waiting for a decision are not reposted.
  5. If there is a delta, write one finding signed with the bot's name: 🎯 what changed · 🔗 one link or n/d · ⭐ score · ➡️ next owner and action. Dense prose, no code fences. Post it to the ops room only.
  6. If there is no delta, stay silent.
  7. Never edit the docs, create bots, or message the owner directly. When a lock arrives in the room (the owner's decision relayed by the fleet director), update the routine prompt's expected count and closed list with a full-prompt update, and log it in memory.
  8. If a run fails or skips, say so in the next stand-up and re-verify before reposting.

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.