External/5chan
1
0
Fork 0
mirror of https://github.com/bitsocialnet/5chan.git synced 2026-08-03 07:41:04 +02:00
Code Issues Packages Projects Releases Wiki Activity

Update AGENTS.md

Browse Source
This commit is contained in:
plebeius
2026-02-18 16:06:36 +08:00
parent e0ed510fc6
commit 91752206f6
1 changed files with 10 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
AGENTS.md
+10 -8
View File
@@ -370,19 +370,21 @@ React Doctor is advisory quality tooling for React architecture/perf/correctness
Never mark a task complete without proving it works. Agents tend to stop too early.
- After code changes, **delegate verification to subagents**: run `yarn build`, `yarn lint`, `yarn type-check` and check the output
- For UI changes, use a subagent to check the browser with playwright-cli
- After code changes, run `yarn build`, `yarn lint`, `yarn type-check` and check the output
- For UI/visual changes, check the browser with playwright-cli to verify rendering
- After touching React components/hooks, run `yarn doctor` and fix any diagnostics
- If verification fails, fix the issue and re-verify — **loop until it passes**, don't leave broken code
- Ask yourself: "Would a staff engineer approve this?"
### Subagent Strategy
### Keep Context Lean
Use subagents to keep the main context window lean.
Delegate heavy work to subprocesses (subagents, background tasks, etc.) to avoid filling the main context window with build logs, test output, and browser snapshots.
- Offload research, exploration, verification, and parallel analysis to subagents
- One focused task per subagent — don't overload a single subagent
- The main agent orchestrates; subagents do the heavy lifting (build logs, test output, browser checks stay out of main context)
- For complex problems, launch multiple subagents in parallel
- One focused task per subprocess — don't overload a single one
- The main agent orchestrates; subprocesses do the heavy lifting
- For complex problems, run multiple subprocesses in parallel
**Note:** These guidelines are universal and tool-agnostic. This is a FOSS project — contributors use different AI tools (Cursor, Claude Code, Codex, etc.) each with their own subprocess/subagent systems. AGENTS.md describes *what* to do (verify, delegate, loop), not *how* your specific tool implements it. If your tool supports custom subagent definitions (e.g. `.cursor/agents/`, `.claude/agents/`), configure them locally — they're gitignored and won't be shared.
### Self-Improvement