agentleFS
aicodekingaicodeking/deepseek-stack/skills/ds-review-loop/SKILL.md

ds-review-loop

Skill80 starsChanged 15 days agoMITView on GitHub

Runs an author and a fresh-context reviewer in a loop on the current change until both agree it is correct, for at most four rounds. Use before or after filing a PR, when the user says review loop, get it reviewed, or /ds-review-loop.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 4 are distinctive, and 1 is 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.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write the goal and out-of-scope list in two sentences.

    line 11
  2. Distinctive:

    Wait for the reply. Do not edit the scoped files while the review runs.

    line 26
  3. Distinctive:

    Keep a ledger: id, severity, outcome, evidence.

    line 28
  4. Distinctive:

    Do not push during the loop.

    line 40
  5. Very specific:

    If you update a PR, add: Reviewed with ds-review-loop: <N> rounds, <reviewer>, verdict <verdict>.

    line 41

≈ “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

Make a change more correct by letting one fresh reviewer push back until you both agree.

Steps

  1. Scope the change. Run git status --short. Do not stash, discard or include unrelated files. Find the base: git fetch origin, then git merge-base HEAD origin/<base branch>. Use one diff command for every round.
  2. Write the goal and out-of-scope list in two sentences. Run the covering checks. Do not send code that fails them.
  3. Pick the reviewer: a stronger model if the harness can reach one (subagent with a model setting, team worker, second CLI), or else a fresh DeepSeek context with no history. Record which one.
  4. Send this brief:

    Review only. Do not edit, commit, push, comment or start subagents.
    Repo: <path>  Base: <sha>  Diff: git diff <sha> [-- <paths>]
    Goal: <goal>  Out of scope: <list>  Checks: <results>
    Read the code around the diff. Find bugs, broken callers and gaps against the goal.
    Flag code the goal does not need. Ignore style unless it hides a bug.
    blocker = goal case is wrong. should = real bug users will hit. nit = anything else.
    Reply: VERDICT: APPROVE or CHANGES, then one line per finding:
    R1 [blocker|should|nit] file:line claim. Fix: suggestion
  5. Wait for the reply. Do not edit the scoped files while the review runs. A timeout or error is not approval.
  6. Judge each finding yourself. Agree: make the smallest fix, preferring to delete or narrow code. Disagree: reply with evidence (file:line, test output). Nit or out of scope: decline in one line.
  7. Keep a ledger: id, severity, outcome, evidence. Re-run checks and git diff --shortstat. If the diff grew, cut it back.
  8. For the next round send the brief, the ledger and the changes since the last round. Ask the reviewer to check only open ids and changed code.

Stop when

  • The reviewer returns APPROVE on the current diff, no blocker or should is open, and you have no concern left.
  • A round returns only nits.
  • Four rounds are done, or one item stays disputed after two exchanges. Report both sides in two or three lines each and let the user decide. This is not approval.

Rules

  • One reviewer at a time. Never agree just to end the loop.
  • Do not push during the loop. Push afterward only if the user asked to push or ship.
  • If you update a PR, add: Reviewed with ds-review-loop: <N> rounds, <reviewer>, verdict <verdict>.
  • Report rounds, reviewer, fixed and declined findings with reasons, and diff size before and after.

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.