Annotate and review coding agent plans and code diffs visually, share with your team, send feedback to agents with one click.

Local-first agent plan annotator

Systematic thinking enforcement plugin for OpenClaw — task tracking, plan mode, hard gates, quality checklists, and audit logging for AI agents

Pi extension for plan mode - read-only exploration and analysis

Don't go into production without these - 3 auto-triggering Claude Code skills for cost and drift prevention. 6 months of practitioner notes.

A Plan-Mode-first Codex skill for extracting project philosophy, anti-goals, and mental harnesses through guided dialogue.

A universal agent skill for structured plan mode. Built natively for OpenClaw and Claude Code. Works with Cursor, Windsurf, Codex CLI, Aider, and more.

A terminal user interface (TUI) for reviewing and iterating on Claude's plans before implementation.

Real-time plan-drift watchdog for Claude Code. Keeps Claude honest to its own plan.

OpenClaw skill: PLAN mode - BUILD mode with auto-debug, auto-test, and Plan Fidelity Score.

Treat your Claude Code plans like database migrations. Auto-capture plan mode into immutable repo history.

Pi extension that adds a persistent Claude/Codex/OpenCode-style plan mode with safe tooling, session-scoped plan files, and inline footer status

V2.1 Alpha — multi-model agent framework with five-layer working memory, plan-mode orchestration, and real quality control.

Force 5 LLMs (Haiku 4.5 + Sonnet 4.6 + Opus + 3 local Ollama) to peer-review every Claude Code plan. Hook-gated. No bypass. Fully local for the open-weight reviewers.

Sharable distillations of working-effectively-with-Claude-Code doctrines + tools (saga-21 corpus)

🛠 Review TypeScript and React code with precise, actionable feedback like a senior engineer for better code quality and faster development.