agentleFS
timmo001timmo001/skills/review-steering/SKILL.md

review-steering

Skill2 starsChanged todayApache-2.0View on GitHub

Review mode for long autonomous work whose design is already settled, with no review requests, sign-offs or design questions. Changes are pushed as they reach a working point, decisions are made from the given design and reported, and the user steers only when they choose to. Use ONLY as an explicitly invoked OpenCode `/goal` objective alongside a loop skill, such as `/goal @issue-loop @inbox-reporting @review-steering`; use `review-stages` instead when each stage needs sign-off.

We split this skill 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. Very specific:

    Don't open code reviews, diff reviews or gated sign-offs, and skip the usual end-of-change review handoff (such as change-review-handoff).

    line 15
  2. Distinctive:

    Don't ask design or direction questions.

    line 16

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 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

Review Steering

The design is already given, in the issues, the goal text and the repository's docs and conventions. Asking the user to review or decide is wasted time for them. They watch the updates and step in with a steer when they want something different; until then, the work never waits on them.

Don't ask

  • Don't open code reviews, diff reviews or gated sign-offs, and skip the usual end-of-change review handoff (such as change-review-handoff). Only open a code review if the goal text asks for one.
  • Don't ask design or direction questions. Where the given design leaves something open, make the call that fits it best, carry on, and name the call in the status update so the user can steer if they disagree.
  • Reach for the user only for what no decision of yours can settle: access, a human-only step, or something the given design rules out or contradicts.

Pushing and updates

Push whenever the branch is at a working point, as the loop describes. With each pushed change, say what changed and where to see it (rendered docs pages, UI, screenshots, previews or URLs, with a link, a command or an attached screenshot). This is information for the user to steer from, not a request for review.

Steering

A steer can arrive at any point, about the current item or work already pushed. Track it as the loop describes: insert it into the todo list at the next good stopping point, keep going with the current focus until it reaches a working point, then pick the steer up. Note it on the issue when it outlasts the session, and don't wait for confirmation. Fix pushed work in new commits rather than rewriting history.

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.