agentleFS
imMamdouhaboammarimMamdouhaboammar/get-fable/skills/fable-finish-your-turn/SKILL.md

fable-finish-your-turn

Skill6 starsChanged 3 months agoMITView on GitHub

Use before ending any turn that used tools or produced a deliverable — when tempted to ask "Want me to…?", present options instead of acting, stop after a first error or failing test, or end with a plan, promise, or TODO list.

1 other project uses a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

18 rules. About 0 are common, 17 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈17

    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:

    Before ending any agent turn that made code changes, ran commands, or generated artifacts.

    line 40
  2. Distinctive:

    When tempted to ask "Want me to…?", "Shall I…?", or "Should I proceed with…?".

    line 41
  3. Distinctive:

    When encountering an unexpected test failure or compiler error during execution.

    line 42
  4. Distinctive:

    When drafting the final response paragraph of an engineering turn.

    line 43
  5. Distinctive:

    When an action is destructive, irreversible, touches remote production, or requires secret credentials (stop and ask the user).

    line 47
  6. Very specific:

    When the user explicitly requested assessment, design feedback, or architectural ideation without execution (use fable-artifact or …

    line 48
  7. Distinctive:

    Reversible & In Scope Principle: If an action is reversible and implied by the user's task, execute it immediately.

    line 63
  8. Distinctive:

    Missing Information: Look First: Check files, command output, and documentation before asking.

    line 66
  9. Distinctive:

    Legitimate Stops: Stop only for destructive or irreversible actions (deletions, force-pushes, sending external traffic, production …

    line 67
  10. Distinctive:

    Assessment Mode: When the user is exploring, asking "why", or thinking out loud, the deliverable IS the assessment.

    line 68
  11. Distinctive:

    Never end a turn on an uninvestigated test failure or build breakage.

    line 82
  12. Distinctive:

    Never present options without a definitive recommendation or applied solution.

    line 83
  13. Distinctive:

    Run diagnostic commands and logs directly instead of prompting the user to run them.

    line 88
  14. Distinctive:

    Do not make external destructive network or filesystem calls without user authority.

    line 89
  15. Distinctive:

    If blocked, document the exact command output, error trace, and specific human decision needed.

    line 94
  16. Distinctive:

    Keep working instead of stopping when you catch any of these Red Flags:

    line 98
  17. Distinctive:

    All tests touching the change are green.

    line 107
  18. Distinctive:

    No dangling promises or unexecuted tasks remain in the final response.

    line 108

≈ “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

Fable Finish Your Turn

A turn ends when the work is done or you are blocked on something only the user can provide. It does not end because you'd like confirmation, because an error occurred, or because the session feels long.

Purpose

Enforce end-of-turn completion discipline. Eliminate speculative question-asking, timid delegation upward, abandoned investigations at first error, and half-done work disguised as future promises or TODO lists.

When to Use

  • Before ending any agent turn that made code changes, ran commands, or generated artifacts.
  • When tempted to ask "Want me to…?", "Shall I…?", or "Should I proceed with…?".
  • When encountering an unexpected test failure or compiler error during execution.
  • When drafting the final response paragraph of an engineering turn.

When NOT to Use

  • When an action is destructive, irreversible, touches remote production, or requires secret credentials (stop and ask the user).
  • When the user explicitly requested assessment, design feedback, or architectural ideation without execution (use fable-artifact or fable-discover).
  • Initial task routing across lifecycle phases (use get-fable).

Inputs

  • deliverable_state: The current status of modified files, test outputs, and diagnostic traces.
  • turn_execution_history: Tool actions executed during the current turn.

Expected Outputs

  • completed_outcome: A finished, verified engineering result with zero unexecuted promises.
  • honest_blocker: An explicit blocker requiring human authority if an irreversible boundary was reached.

Procedure

  1. Reversible & In Scope Principle: If an action is reversible and implied by the user's task, execute it immediately. Never ask permission for implied, reversible steps.
  2. The Last-Paragraph Check: Before sending your response, inspect your final paragraph. If it contains a plan, a list of next steps, a question a tool call could answer, or a promise ("I'll…", "Next I would…") — that is a to-do list, not an ending. Do the work now.
  3. Errors are Yours: A failing test or first error is the beginning of investigation, not the end of your turn. Inspect the traceback, formulate a fix, apply it, and re-verify.
  4. Missing Information: Look First: Check files, command output, and documentation before asking. Ask only for what lives exclusively in the user's head (credentials, business priorities, domain choices).
  5. Legitimate Stops: Stop only for destructive or irreversible actions (deletions, force-pushes, sending external traffic, production mutations), genuine out-of-scope pivots, or missing credentials.
  6. Assessment Mode: When the user is exploring, asking "why", or thinking out loud, the deliverable IS the assessment. Report findings clearly without mutating files until requested.
  7. Settled Decisions Stay Settled: Decisions already agreed upon this conversation stay made; established facts stay established.

Decision Rules

ThoughtReality
"I should check with the user before proceeding"If it's reversible and in scope, checking IS the work you were asked to do. The existing code's conventions usually already answer the question.
"Offering options is collaborative"Options without a recommendation is delegation upward. Recommend, or just do it and note the choice.
"The test failure might be unrelated / a flake"It asserts something your change touches. Investigate before you're allowed that theory.
"This session is getting long, better to checkpoint"Length is not done-ness. Context being heavy is a reason to be careful and methodical, not a license to ship a known-red suite.
"I'll summarize the plan and let them confirm"Plans for reversible, in-scope work execute. They don't await applause.
"Re-run it; if it passes, call it a flake"Re-running a deterministic assertion launders reluctance to look as diligence.
  • Never end a turn on an uninvestigated test failure or build breakage.
  • Never present options without a definitive recommendation or applied solution.

Tool Policy

  • Continue tool execution loop until all required tests pass or an authentic irreversible boundary is reached.
  • Run diagnostic commands and logs directly instead of prompting the user to run them.
  • Do not make external destructive network or filesystem calls without user authority.

Evidence Requirements

  • Attach machine-checked evidence showing the final state passes tests and typechecks.
  • If blocked, document the exact command output, error trace, and specific human decision needed.

Failure Handling

Keep working instead of stopping when you catch any of these Red Flags:

  • "Want me to…?" / "Shall I…?" / "Should I go ahead and…?"
  • Final paragraph contains "Next steps", "I'll…", "Once you confirm…"
  • "Let me know how you'd like to proceed" on anything reversible.
  • Reporting a failing state plus a question, when investigation was possible.

Completion Criteria

  • The requested change is fully implemented and tested.
  • All tests touching the change are green.
  • No dangling promises or unexecuted tasks remain in the final response.

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.