Bootstrap CLAUDE.md, .claude/agents/, hooks, and plugin wiring with /agent-init
Cost-unrestricted patterns: /visual-qa (visual regression + LLM analysis via Playwright MCP) and /wiki — a self-auditing project knowledge base in .wiki/ (Karpathy LLM-Wiki pattern): INDEX.md router, provenance-graded pages (A/B/C), contradiction preservation, and a 'compile' self-audit gate (diff=0); SessionStart digest prints wiki status on open
Run agent-init for Codex CLI projects — emits AGENTS.md + .codex/skills/
Run agent-init for GitHub Copilot CLI — emits .github/copilot-instructions.md + AGENTS.md
Run agent-init for Gemini CLI (a.k.a. 'antigravity') — emits GEMINI.md + .gemini/skills/
Run agent-init for Cursor — emits .cursor/rules + .cursor/agents/
visual-qa scaffold for Codex CLI — emits .visual-qa.json + Playwright MCP entry; ships near-native /wiki (self-auditing knowledge base in .wiki/), the closest Codex port of the harness-floor wiki skill
Theme B — cost-conscious long-session optimisation: context-mode integration, prompt cache priming, summariser hooks, end-of-session audit
Theme B port for Cursor — advisory-only thrift via .cursor/rules/thrift.mdc
Theme B port for GitHub Copilot CLI — file-backed thrift state + official .github/hooks/ patches
Theme B port for Codex CLI — TOML-aware ~/.codex/config.toml patcher + OpenAI rate table
Theme E port for Codex CLI — prompt-level systematic debugging with structured error parsing, .debug-state.json persistence, and git/input bisection
Theme B port for Gemini CLI — Vertex AI rate table with min-token + storage-hour terms
Theme D — codebase exploration: parallel-dispatch tree-walker + dependency graph, cached by git HEAD, /explore where + /explore deps queries
Theme E — systematic debugging: reproduce → isolate → hypothesize → verify workflow with .debug-state.json checkpointing + 10-format error parser + bisection lib
Data analysis harness for notebook execution, SQL validation, artifact diffs, and durable verification evidence