agentleFS
UnfoldCRO-OrganizationUnfoldCRO-Organization/cro-agent-skills/skills/rea-research/SKILL.md

rea-research

Skill0 starsChanged 22 days agoMITView on GitHub

Evaluate REA for authorized reverse engineering of binaries, Electron apps, closed-source JavaScript, Android packages or observed runtime behavior. Use for reverse-engineering research when source is unavailable. Do not install automatically.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    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
  1. Distinctive:

    Use this reference only when ordinary source inspection and existing debugging tools cannot answer the question, and the target belongs to …

    line 14
  2. Distinctive:

    Do not invoke for ordinary Shopify Liquid or theme tasks, public-site SEO/CRO audits, open-source repository understanding, screenshots …

    line 22
  3. Distinctive:

    Research first: Read the upstream README, SECURITY.md, DISCLAIMER.md, and relevant workflow instructions.

    line 26
  4. Very specific:

    Not an installation instruction: Do not run npx, npm install, rea setup, rea update, scripts, or MCP registration merely because this skill …

    line 27
  5. Distinctive:

    Least privilege: Use an isolated development environment and disposable browser profile without client logins or unrelated secrets.

    line 28
  6. Very specific:

    Pin releases: When approved, use an inspected and pinned release.

    line 31
  7. Distinctive:

    No automatic installations, runtime observation, client access, or production actions.

    line 42

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

REA research (on demand)

Official upstream: https://github.com/morluto/rea Official project: https://rea.tools/ Package: rea-agents (verify the actual release and checksum before installation).

Routing gate

Use this reference only when ordinary source inspection and existing debugging tools cannot answer the question, and the target belongs to the agency or the client has authorized the particular analysis.

Appropriate cases:

  • Inspect a supplied Electron application or JavaScript bundle with no usable source.
  • Investigate documented runtime behavior in a user-owned, isolated browser.
  • Inspect Android packages or native binaries when the project genuinely requires it and analysis is authorized.
  • Compare permitted behavior across versions to understand interoperability or regressions.

Do not invoke for ordinary Shopify Liquid or theme tasks, public-site SEO/CRO audits, open-source repository understanding, screenshots alone, or routine Chrome DevTools diagnostics. Prefer the existing seo-tech-stack, ecom-cro, shopify-ui-audit, source inspection and browser tools.

Safety and approval

  1. Research first. Read the upstream README, SECURITY.md, DISCLAIMER.md, and relevant workflow instructions. Use the official repository as the source of installation instructions, not an arbitrary copied skill.
  2. Not an installation instruction. Do not run npx, npm install, rea setup, rea update, scripts, or MCP registration merely because this skill was loaded. Get explicit user approval for installation or configuration changes.
  3. Least privilege. Use an isolated development environment and disposable browser profile without client logins or unrelated secrets. Never use production credentials or private third-party binaries unless separately authorized.
  4. Authorization per target. Being able to access a site or binary is not permission to bypass authentication, DRM, licensing, anti-bot controls or to copy protected code/assets.
  5. Separate permission for active actions. A read-only artifact review does not authorize program execution, browser actions, attaching to a process, device changes, interception, network probing, or production changes. Explain impact and obtain approval.
  6. Pin releases. When approved, use an inspected and pinned release; avoid unattended @latest installation and auto-updates. Review dependencies and setup dry-run before changes.
  7. Evidence and uncertainty. Report observed artifacts, versions, timestamps, limitations and uncertainty. Do not present reverse-engineered inferences as source-verified facts.
  8. Do not import code. Implement independently using functional requirements and original assets, subject to client ownership and applicable licenses.

Decision output

Return one of:

  • Not needed: existing source/browser tooling can complete the task.
  • Useful, no installation: provide investigation plan and requested evidence.
  • Install proposed, approval required: show exact package version, installation scope, configuration changes, credentials/process/browser exposure, rollback plan and prerequisites.

No automatic installations, runtime observation, client access, or production actions.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.