tighten-file
Skill4 starsChanged 7 months agoMITView on GitHub
File-level tightening pass on an instruction file (CLAUDE.md, skill, agent prompt, style guide) using `tighten-instruction` as the lens. TRIGGER when: user says 'tighten/simplify this file/skill/CLAUDE.md', 'cut this down', 'titan instruction on this file'; user points at a verbose instruction file and wants it leaner.
In this repository42 other files it gives its agents.
- AGENTS.mdAGENTS.md
- agent-soulSkill
- audit-transcripts-for-learningsSkill
- best-answerSkill
- check-coherenceSkill
- compress-fileSkill
- deep-modulesSkill
- design-critiqueSkill
- durable-docs-updateSkill
- execute-chatSkill
- execute-planSkill
- explain-deeplySkill
- explain-simplySkill
- find-gapsSkill
- fix-verify-loopSkill
- generate-mockupsSkill
- grill-meSkill
- handoffSkill
- interview-meSkill
- jtbdSkill
- map-design-languageSkill
- memory-pruneSkill
- multi-agent-analysisSkill
- place-factSkill
- post-purchase-ui-extensionSkill
- product-interviewSkill
- prove-behaviorSkill
- refine-fileSkill
- second-opinionSkill
- sentry-analysisSkill
- shopify-dev-mcpSkill
- structure-proseSkill
- tech-designSkill
- test-completed-planSkill
- tighten-instructionSkill
- tldrSkill
- triageSkill
- trim-specSkill
- two-pass-reviewSkill
- validate-answerSkill
- vet-factSkill
- write-planSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 4 are distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 29
Keep load-bearing rationale.
- Very specific: line 46
Write Nothing to tighten — already clear.
- Distinctive: line 52
every Step 1 invariant survived.
- Distinctive: line 53
every exact token and anchor remains valid.
- Distinctive: line 54
every cross-reference still resolves.
- Distinctive: line 55
no rationale lost its constraint.
≈ “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.
Files
Tighten File
Apply tighten-instruction at three levels: whole file, section, instruction.
Steps
Step 0 — Load
Invoke the tighten-instruction skill via the Skill tool. This file-level workflow owns the apply gate.
Step 1 — Pin meaning
State the file's purpose in one line. List every distinct instruction or fact that must survive, including exact tokens, anchors, output contracts, and load-bearing rationale.
Step 2 — Find and score edits
Apply the loaded lens at whole-file, section, and instruction level. For each independent edit, record:
- the current text;
- the proposed text, or
cut entirely; - the level: whole-file, section, or instruction;
- confidence
0.00–1.00that the edit improves clarity without changing meaning.
Keep load-bearing rationale. When a reason is the non-derivable fact, shape it as behaviour — constraint; do not cut it as explanation.
Step 3 — Gate and apply
Show the plan, then:
- apply every edit at
c ≥ 0.80; - hold lower-confidence proposals and leave their text unchanged;
- apply whole-file and section edits before instruction edits;
- drop any later proposal an applied edit dissolves.
**Tightening plan — <file>:**
- [0.92] section: <current> → <proposed>
- [0.61] instruction: <current> → <proposed> · held (< 0.80)Write Nothing to tighten — already clear. when no edit qualifies.
Step 4 — Prove and report
Re-read the changed file cold. Confirm:
- every Step 1 invariant survived;
- every exact token and anchor remains valid;
- every cross-reference still resolves;
- no rationale lost its constraint.
Revert or fix any edit that fails this proof.
Report applied edits, held proposals, dissolved proposals, and net clauses, lines, and words removed.
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
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.

