agentleFS
IcodeNetIcodeNet/agent-skills/skills/feedback-checkpoint/SKILL.md

feedback-checkpoint

Skill0 starsChanged 4 months agoMITView on GitHub

Status checkpoint to unstick a run: detect loops and stuck states, report Done / Next / Blocking, and recover one step at a time. Use when the user asks for a status update, says "checkpoint", or when you suspect you are looping or stuck.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are 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.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run this to take stock and unstick a session.

    line 12
  2. Distinctive:

    Do only that one step, then stop and wait for user feedback.

    line 24
  3. Distinctive:

    Keep the task list up to date.

    line 25
  4. Distinctive:

    Stop and wait for user direction (e.g.

    line 29

≈ “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

Feedback Checkpoint

Run this to take stock and unstick a session.

  1. Check for a loop. Have you done 3+ similar actions in a row, repeated the same failing step, or kept re-reading/re-editing the same area without progress? If yes, treat as loop.
  2. Check for stuck. Are you blocked (error, missing type, unclear requirement) or have you spent more than 2–3 turns on the same task without finishing? If yes, treat as possibly stuck.
  3. Respond in one of these ways:

    If in a loop:

    • Say you may be in a loop, what you were trying to do, and the single next step you will take.
    • Do only that one step, then stop and wait for user feedback.
    • Keep the task list up to date.

    If stuck:

    • Output a 3-bullet status: Done, Next, Blocking.
    • Stop and wait for user direction (e.g. "Proceed", "Simplify scope", or clarification). Do not start another round of tool calls until the user responds.

    If neither:

    • Give a short 3-bullet status anyway (Done / Next / Blocking) so there is a clear checkpoint.
  4. Rule: one task at a time when recovering. Complete one step, then rely on the next turn or user confirmation before continuing.

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.