Machine-local half of a Claude Code setup: global rules, skills, hooks, plugin manifests, per-project memory. No credentials.
Resources
2Install
npx skillscat add surreptakos/claude-dotfiles/aac-skills-research Install via the SkillsCat registry.
SKILL.md
Spin 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. When a source is a large dataset or dump rather than prose, write and run code against it to pull out the relevant facts rather than reading the raw data into context.
- Write the findings to a single Markdown file, citing each claim's source.
- Save it where the repo already keeps such notes; match the existing convention, and if there is none, put it somewhere sensible and say where.
- Before calling the note finished, ground-check it where the repo carries the checker (claude-dotfiles does):
node tools/check-evidence.js <note.md>. Every number, date and quote must appear verbatim in a source the note cites — a Markdown link or aSource:line; a path named in backticks is a mention, not a citation. Exit 1 names the note line whose fact is missing from the source; exit 3 means a source could not be read, which is not a pass. Fix the note or cite the source the fact actually came from.