skills.sh
AvailableUniversal skill installation, including Claude Code and Codex.
npx skills add https://github.com/ulpi-io/skills-autonomous-engineering --skill auto-performanceauto-performance
Keeps only measured performance wins that exceed variance without changing correctness.
WORKFLOW / 05/auto-performance [target — endpoint/function/page/metric, or a perf budget] (default: profile to find it)
- 01
Define the metric, target, benchmark command, and variance-aware baseline.
- 02
Profile representative work to locate the real hotspot.
- 03
Change one hotspot and rerun the same benchmark.
- 04
Keep the change only when the gain is real and correctness remains green.
- 05
Stop at the target or diminishing returns and report every reverted attempt.
▸ 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
- A measurable performance target or suspected regression exists.
- Profiling should identify the real hotspot before optimization.
- Candidate optimizations need apples-to-apples benchmark evidence.
Do not use when
- No reproducible metric or benchmark can be established.
- The problem is functional correctness or a new feature.
- The goal is readability rather than runtime performance.
▸ output contract
What must be true when it stops.
Produces
- Metric and variance-aware before/after benchmark
- Kept optimizations with per-hotspot measured gains
- Tried-and-reverted optimization ledger
- Target-met or stalled verdict
Guarantees
- No optimization starts without a reproducible baseline.
- Profiling evidence chooses the target instead of intuition.
- Acceptance requires a gain beyond run-to-run variance.
- Tests and adversarial regression checks protect correctness.
- Unmeasured or noisy claims are reverted, never rounded into wins.
Honest failure states
- no_baseline — the benchmark cannot be made reproducible
- reverted_no_gain — the delta is inside measurement noise
- reverted_regression — correctness changed
- diminishing_returns — remaining gain is not worth added complexity
▸ 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-performance 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-performance.
▸ composes with