Open-source repos
REA: Reverse Engineer Anything
Explain software behavior from shipped artifacts and runtime evidence
This page provides source information and editorial guidance. Recommendations and trust labels do not mean every skill here was tested. Results for individual skills are on the skill data pages. See skill data · Listing policy
Overview
REA’s reverse-engineer-anything skill routes a target to binary, package, Electron/JavaScript, Android or recorded-runtime analysis, then ties conclusions to evidence identifiers. It suits investigations of shipped software, version comparisons and reconstruction. Ordinary architecture analysis of a complete source repository belongs with standard repository tools.
SkillFlux sample check
On 2026-10-31, SkillFlux pinned this source at commit d48e472, drew a seeded sample of 5 from the 5 entries it contains, and ran static health check 1.0.0 on each. 40% had no failing check, said when to use the skill and carried a license.
- 5/5sampled entries had a SKILL.md we could check
- 0/5fail at least one check
- 3/5do not say when to use the skill
- 0/5have no license we could find
- 0/5bundle scripts the agent may run
- 0/5include scripts that call the network
- 0/5include destructive shell commands
- 1/5exceed the 5,000-token body guideline
See the 5 sampled entries
| Entry | Result | Flags |
|---|---|---|
| morluto/rea/.agents/skills/rea-tool-design | 14 pass, 1 warn, 0 fail | — |
| morluto/rea/.agents/skills/reverse-engineer-anything | 12 pass, 3 warn, 0 fail | — |
| morluto/rea/.agents/skills/rea-changelog-update | 14 pass, 1 warn, 0 fail | — |
| morluto/rea/.agents/skills/rea-ci-tests-audit | 15 pass, 0 warn, 0 fail | — |
| morluto/rea/.agents/skills/rea-analysis-optimization | 15 pass, 0 warn, 0 fail | — |
The sample reads files only; it does not run any skill and does not describe entries outside the sample. Flags describe what a skill contains; they are not verdicts. Raw data · Method
What the skills help you do
Inputs, methods and outputs described in the collected source version.
Route by target type
ASAR or JavaScript trees, managed assemblies, Android packages and native binaries use different entry points, avoiding native decompiler setup for a static JavaScript task.
Read the source instructions ↗Build evidence around investigation questions
Turn cross-feature questions into a checklist, start with the smallest useful overview, and trace entry points, data and state changes while recording evidence and coverage gaps.
Read the source instructions ↗Bound version-comparison claims
Artifact, function and evidence-set comparisons retain their scope. Static changes are candidate behavioral clues, while missing or incomplete coverage remains unknown.
Read the source instructions ↗Reconcile static clues with passive observations
Collect passive evidence from an already-open browser or Electron target and reconcile it with static application evidence. Seeing a script does not prove every module executed.
Read the source instructions ↗A typical workflow
Identify the artifact and question
Provide a specific local artifact, target application or existing capture and list the behavior questions to answer.
Choose an available analysis route
Choose tools by target type. Diagnose configuration only when the connection is absent or stale, and reuse matching evidence.
Investigate specific evidence gaps
Read summaries and evidence first, then make focused queries or runtime observations only for questions that remain unanswered.
Deliver findings with explicit limits
Mark each question answered, partially answered or unresolved; link evidence identifiers and distinguish observations, inferences and unknowns.
Best for
- Developers with a package or binary who need to explain a feature or determine what changed between versions.
- Investigation and compatibility work that needs to distinguish static clues, observed behavior and unresolved questions.
Use cases
- Coding & development
- Security & review
- Tools & MCP
SkillFlux notes
This page analyzes target routing and evidence workflows from the skill entry, five local guides and MIT license at a pinned snapshot; it does not claim those targets were tested. REA’s MCP server, CLI and analysis engines are separate runtime dependencies to prepare for your platform.
Installation and use
Open the original repository and read its README, the selected skill instructions and any linked scripts. Check installation paths, host requirements, dependencies and licenses, then follow the project documentation to verify a small task in a workspace you can restore.
Limitations
- The skill files provide instructions; they do not register REA MCP or install analyzers. Native, Android, firmware and browser targets have different dependencies, and the connected server’s actual tools determine availability.
- Static differences do not prove runtime causality, and passive browser or Electron observation does not mean clicking or executing scripts. Captures that execute a target need an explicit target and scope.
Alternatives
- MCP Servers
Use MCP Servers to discover alternative MCP integrations when the analysis tool is not yet chosen; it does not replace REA’s evidence workflow.
- Superpowers (obra)
Compare Superpowers for development, debugging and testing when complete project source is available; REA focuses on artifacts and runtime evidence.