CommunityWriting & Editinggithub.com

bwkw/dotagents

Personal AI development toolkit — skills, hooks, and verification gates that work in both Claude Code and Cursor. Installed once globally; product repositories stay untouched.

What is dotagents?

dotagents is a Claude Code agent skill that personal AI development toolkit — skills, hooks, and verification gates that work in both Claude Code and Cursor. Installed once globally; product repositories stay untouched.

Works with✓Claude Code~Codex CLI✓Cursor
npx skills add bwkw/dotagents

Installed? Explore more Writing & Editing skills: steipete/notion, langchain-ai/langchain, bytedance/podcast-generation · View all 6 →

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

/skill-name — one-line role

Important: this skill never modifies source code. (Investigation and review skills only. Delete this line for skills that do write code, and say plainly what they are allowed to touch.)

Preconditions

Stop immediately if any row fails. Report which condition failed. Do not continue.

ConditionIf unmet
…Stop, report the unmet condition, do not proceed

Position in the workflow

UpstreamThis skillDownstream
…/skill-name…

Files to read

Always read

FileWhy
……

Read only if

FileTrigger condition
……

Reading everything "just in case" is forbidden. Load the conditional table only when its trigger actually applies.

Steps

Step 1. …

Evidence discipline

  • Report only what you verified directly.
  • Cite locations as path/to/file.ts:L42.
  • When you have no basis for a claim, write "could not confirm" — do not guess.
  • Separate fact from inference explicitly.
  • Attach a URL to any external claim.

Output

Where it goes and what shape it takes. If the template runs long, move it to ${CLAUDE_SKILL_DIR}/reference/output-template.md and point at it here.

Done when

  • …

Next

What to run next, and whether to /clear first.

Related Skills