skills.sh
AvailableUniversal skill installation, including Claude Code and Codex.
npx skills add https://github.com/ulpi-io/skills-autonomous-engineering --skill auto-simplifyauto-simplify
Reduces reader cost one small edit at a time while proving observable behavior did not change.
WORKFLOW / 05/auto-simplify [scope — a diff, path, or module] (default: the current diff)
- 01
Set the scope, capture the diff, and establish a green behavior baseline.
- 02
Rank concrete simplification opportunities by reader cost and risk.
- 03
Make one smallest clarifying edit.
- 04
Run relevant tests and an adversarial semantic comparison; revert on doubt.
- 05
Repeat until a full round is dry, then report what stayed and what was kept on purpose.
▸ operating envelope
Use it deliberately.
The shortest useful definition of where this skill helps—and where it should stay out of the way.
Use when
- Working code contains duplication, dead code, needless branching, or over-abstraction.
- A recent diff is correct but harder to understand than necessary.
- You want a behavior-preserving cleanup with explicit proof.
Do not use when
- The intended change alters behavior, fixes a defect, or adds a feature.
- The goal is runtime performance; use auto-performance.
- The purpose of the code is unknown and cannot be established.
▸ output contract
What must be true when it stops.
Produces
- Behavior-preserving simplification diff
- Ledger of kept, investigated, and reverted opportunities
- Final relevant-suite result
Guarantees
- Every retained edit passes tests and an adversarial behavior-preservation check.
- Code whose purpose is not understood remains behind Chesterton’s Fence.
- Edits remain small enough to attribute and revert independently.
- Clarity wins over line count or cleverness.
- A non-dry or stalled run is reported honestly.
Honest failure states
- reverted — a candidate changes behavior or loses the proof signal
- fenced — the code purpose cannot be established safely
- stalled — worthwhile opportunities remain but no verified progress is possible
▸ enforcement
Know which claims are executable.
Scripts are linked when this skill owns deterministic machinery. The remaining rules are explicit operating contracts and must not be marketed as hooks.
Procedural contract
This skill composes platform tools and verification evidence; it does not claim a dedicated guard script.
▸ install and invoke
One skill, three honest surfaces.
Claude Code plugin
AvailableInstall the full plugin, then invoke /auto-simplify or let routing select it from context.
Codex plugin
Work in progressA work in progress on the codex-native-plugin branch — build it, start a new session, then invoke $autonomous-engineering:auto-simplify.
▸ composes with