agentleFS
Surething-ioSurething-io/cockpit/skills/go/SKILL.md

go

Skill44 starsChanged 8 months agoMITView on GitHub

Implement agreed work in self-verifying slices without stopping, logging every decision the spec did not cover.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are 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.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  • could be automatic: a setting could enforce it
  1. Distinctive:

    Stop and ask only for: missing information that makes progress impossible, a destructive or irreversible operation (deleting data, …

    line 32

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Implementation mode: build what was agreed, in small slices that each prove themselves, without stopping for sign-off.

input  = the path or brief given, else what this conversation agreed (a PRD, an analysis, a plan)
accept = its success criteria; if none can be inferred, ask once before starting
notes  = {{COCKPIT_DIR}}/skills/go/notes/<basename of cwd>-<feature>.md
                                                 # same task continued → append under a new round heading

while slices remain:
    pick the next smallest slice that runs and can be checked on its own
    implement it                                   # minimum change, KISS
    verify by running it: commands, requests, tests — not by reading the code
    if it fails: fix and re-verify, at most 3 tries
                 then mark it blocked, note why, continue with what does not depend on it
    for each choice the spec did not cover:
        cheap to undo  → take the simplest option, log it to notes
        costly to undo → stop and ask
    emit: Stage N — <what> — <how verified> — <result>

emit recap:
    ## End-to-end   the full user flow, run once: steps and result
    ## Decisions    the notes file as a markdown link: [<basename of cwd>-<feature>.md](<absolute path>)
    ## Residuals    what is left, blocked or known broken, with suggested priority

Stop and ask only for: missing information that makes progress impossible, a destructive or irreversible operation (deleting data, force-pushing, rewriting history), or a choice that is costly to undo. Never stop just to report a finished slice.

Notes — one Y-statement line per decision: <chose X over Y> — <to achieve …>, accepting <cost>. Also log spec items you had to change and anything the user should know before shipping. It is not a work log: slices finishing, tests passing and builds going green stay out. It lives outside the repo, and Cockpit deletes it after 30 days untouched.

Good run: after reading the recap and the notes file, nothing in the code surprises the user.

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.