REA Analysis Optimization
Guides an agent to find and fix wasteful work in REA binary-analysis tools, verified by before/after checks.
- Repository
- morluto/rea
- Stars
- 420,737
- Licence
- MIT
- Files
- 3 files
What it does
Once installed, the agent approaches performance work on REA's analysis providers in a disciplined way: it traces a real task to the step doing unnecessary work, changes that step at its owner, and checks that results, coverage and evidence stay equivalent. It also requires an independent correctness check, careful cache lifetime and invalidation design, and honest reporting of measurements, including the option to leave code alone when a benchmark shows no real gain. It does not itself grant permission to run analysis targets, publish or merge changes.
Great for
- Maintainers of the REA project working on analysis speed, memory use or cache design
- Contributors investigating a concrete latency, CPU or memory bottleneck in a provider or shared workflow
- Developers who need a structured method for judging whether a performance change is safe
Skip it if
- Routine code cleanup or refactoring, which the skill says should use correctness checks instead of synthetic benchmarks
- CI or test-suite speed issues, which the skill assigns to a different skill (rea-ci-tests-audit)
- General users of REA who only want to analyse a binary; the skill is about changing REA's implementation and could lead to risky edits if applied carelessly
Try asking
- “Investigate why repeated queries against the same binary are slow and propose a cache with defined invalidation”
- “Compare before and after timings for a shared decoding change, including small inputs and peak memory”
- “Review whether a proposed index or batch endpoint actually removes enough work to justify its maintenance cost”
- “Decide whether a token-usage reduction in REA's tool outputs is measurable and worth keeping”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 3 original files, the MIT licence and an attribution note.
rea-analysis-optimization.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip rea-analysis-optimization.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip rea-analysis-optimization.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Give the agent a concrete bottleneck or regression to target; the skill is designed to reject speculative optimization.
- Ask it to use a separate checkout for experimental changes so other work is not disturbed.
- Read the referenced Kuna case study for context, but treat its numbers as pinned past observations, not current guarantees.
What's inside
- agents/openai.yaml
- references/kuna-case-study.md
- SKILL.md
Redistributed unmodified from morluto/rea under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.