agentsclimarketplace
Sign in

Research

Skill rauleburro/skills/skills/research

Conduct focused technical research on a problem before implementation. Use when the user asks to investigate, research, deep-dive, or understand a technical issue (bugs, platform quirks, integration problems). Produces a concise consolidated report plus optional detail docs.From its SKILL.md

Install
npx -y skills add rauleburro/skills --skill research

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.1 KB, 453 tokens by cl100k_base, as published. Nobody here has run it

Research Skill

Investigate a technical problem efficiently and persist what you learn so the team never re-investigates the same topic twice.

1. Scope the topic

Extract from the user's message (ask only if genuinely unclear):

  • Topic slug: kebab-case folder name, e.g. bluetooth-audio-routing
  • Stack: platform/language/framework
  • Symptom: observed vs expected behavior
  • Known: hypotheses already tried

2. Prepare the workspace

Run scripts/init-research.sh {topic-slug} to create the docs folder.

3. Launch 3 default agents in parallel

Each agent gets a one-line angle plus the shared context. Read reference/agent-prompts.md when you need the exact prompt wording.

  1. Codebase analysis (Explore agent) — find relevant code, gaps, file:line references.
  2. Root-cause theory (web agent) — why it happens at the OS/framework/protocol level.
  3. Solutions & validation (web agent) — native APIs, libraries, community-confirmed fixes.

Spawn the optional 4th/5th agent only if the user explicitly asks for library comparison or cross-platform API deep-dive.

4. Synthesize

Create exactly these two files:

  • docs/{topic}/NOTAS.md — raw findings, snippets, URLs, quick observations.
  • docs/{topic}/RESEARCH_CONSOLIDADO.md — structured master doc.

Use reference/consolidated-template.md only when you need the full template. Optional detail files (03_android_api.md, 04_librerias.md, etc.) are created only when findings justify them.

5. Finalize

Run scripts/finalize-research.sh {topic-slug} to validate the master doc exists and append an iteration row. Then report to the user:

  1. Root cause (with file:line if applicable)
  2. Top 3 non-obvious findings
  3. Files created
  4. Suggested next step

Keep the chat report concise; the detail lives in the docs.

What ships with it: 4 files

4.0 KB alongside SKILL.md, 2 of them executable

reference/

scripts/

Keep looking

Skills are one crate of 326,645. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.