About research
A skill for delegating research to a background agent that checks primary sources and saves cited findings as a Markdown file in the repository. It suits coding agents that need documented research without blocking ongoing work.
Key features
- 01Background-agent research
- 02Primary-source verification
- 03Claim-level citations
- 04Repository-local Markdown findings
- 05Completion notification with file path
Use cases
- 01Researching a question while other work continues
- 02Tracing claims to official documentation or source code
- 03Saving cited findings alongside repository work
Install
Copy the command and run it in a terminal.
$
npx skills add https:/After installing, type/researchin Claude Code to confirm it loaded.
Bundled filesagents/openai.yaml
Details
From the project READMESpin up a background agent to do the research, so you keep working while it reads.
Its job:
- Investigate the question against primary sources (official docs, source code, specs, first-party APIs), not a secondary write-up of them. Follow every claim back to the source that owns it.
- Write the findings to a single Markdown file, citing each claim's source.
- Save it inside the repo's working tree, where the repo already keeps such notes: match the existing convention, and if there is none, pick a repo-relative location (e.g.
docs/research-<slug>.mdor next to the topic it concerns) and state it. The findings path must resolve inside the repo — check withgit rev-parse --show-toplevel— so the file is a committed artifact of the work it informs.
The agent's completion is pushed back to you: when it finishes (or fails, or is stopped), a notification arrives in your context carrying the findings path. Read the file when the push lands — don't poll for it, and don't plan around reading it at a fixed time.
Comments
Markdown supported