agent-flywheel-plugin

Relentless goal/design interview that sharpens framing and writes durable docs (brainstorm artifact, ADRs, glossary) as decisions crystallize. Use when refining a flywheel goal, pressure-testing scope, or "grill with docs".

Start or resume the full agentic coding flywheel. Drives the complete workflow: scan → discover → plan → implement → review.

Set up flywheel prerequisites for this project.

One-shot diagnostic of every flywheel dependency — MCP connectivity, Agent Mail liveness, br/bv/ntm/cm binaries, node version, git status, dist-drift, orphaned worktrees, and checkpoint validity. Use when debugging toolchain issues, before starting a new session, after /flywheel-cleanup, or as a CI gate.

Strategic gap analysis between vision (AGENTS.md / README.md / plan docs) and what's actually implemented. Converts gaps into beads, optionally launches a swarm. Use when "reality check", "where are we really", "gap analysis", "did we drift", or before declaring a long-running project done.

Launch a parallel swarm of agents to implement multiple beads simultaneously.

Run a state-aware /dueling-idea-wizards duel inside the flywheel. Routes artifacts into discovery, plan, or review pipelines based on the current phase. Use when the user invokes /flywheel-duel directly, or when start.md routes here from a Discover/Plan/Review menu.

Full health check of the codebase and flywheel dependencies.

Check the status of running swarm agents and bead progress.

Mine cass session history for recurring workflows and propose them as new skills, refinements to existing skills, or flywheel pipeline tweaks. Use when "what should we automate next?", "find repeated rituals", or after a long stretch of agent work to harvest tacit knowledge into beads.

Generate and operationalize improvement ideas for projects. Use when brainstorming features, planning improvements, creating beads from ideas, or "what should we build next".

Check if the codebase has drifted from the implementation plan.

Sweep docs/solutions/ to consolidate, replace, or archive stale and duplicate learnings — never auto-deletes.

One-sentence summary of what this skill does. Start with a verb, keep it specific enough that Claude can decide when to invoke it.

Deep research on an external GitHub repository to extract implementation insights.

Iterative UI/UX polishing for Stripe-level visual quality. Use when app already works and looks decent, wanting to elevate to world-class through multiple passes. Not for complete overhauls.

Run a structured UI/UX polish pass on implemented code. Spawns a scrutiny agent to generate 15-30 improvement suggestions, presents them numbered for user selection, converts selected ones to beads, and triggers implementation. Use after initial implementation is complete.

Remove telltale signs of AI-generated "slop" writing from documentation. Use when polishing README files, API docs, or any public-facing text to sound authentically human.

Stop the current flywheel session and reset state.

Audit the codebase for bugs, security issues, test gaps, and dead code.

Clean up orphaned git worktrees from crashed or stopped sessions.

Fast path to apply a targeted fix without running the full flywheel.

Refine a specific agent skill using session evidence and feedback.

Review and improve all loaded agent skills based on session patterns and feedback.

Roll back a completed bead implementation.

Targeted scan of specific paths or concerns without a full flywheel scan.

Show current flywheel status, bead progress, and inbox messages.

Stop all running swarm agents and release their file reservations.

Submit structured feedback about a tool or command to improve future flywheel sessions.

Search, store, or manage CASS long-term memory.

Manage AI coding CLI accounts with sub-100ms switching. Use when hitting rate limits on Claude Max, GPT Pro, or Gemini Ultra subscriptions and need instant account swapping without browser OAuth.

Mine past agent sessions for working prompts, decisions, and patterns. Use when "what did I ask?", "find that prompt", session archaeology, or agent history.

Rebuild CHANGELOG.md files and release histories from git, tags, releases, and issue trackers. Use when writing changelogs, version timelines, or agent-facing project history summaries.

Cross-validate decisions using multiple AI models (Codex, Gemini, Grok). Use when "get a second opinion", evaluating approaches, or high-stakes decisions.

Comprehensive code review workflow using Ultimate Bug Scanner (UBS). Use when reviewing code, scanning for bugs, validating AI-generated code, or running pre-commit quality checks.