sabu

Reconcile multiple plans for the same piece of work — written by different models or sessions (e.g. rubber-duck runs in Fable, Opus, Kimi, GPT) — into one final plan, finding where they agree and diverge, optionally convening a council of lens-based reviewers to judge them. Use when the user wants to compare, reconcile, or merge multiple plans, asks for a council, verdict, or second opinion across plans, mentions plan files like <model>-<topic>.md, or says they ran the same planning exercise in multiple sessions — even if they don't use the word "reconcile".

Interview the user relentlessly about a plan or design, one question at a time, until reaching shared understanding — resolving each branch of the decision tree. When the project documents its domain (CONTEXT.md, ADRs), challenge the plan against that language and capture decisions inline. Use when the user wants to stress-test a plan or an issue, asks your opinion on a plan or design, or mentions "rubber-duck" / "rubberducking".

Disciplined diagnosis loop for hard bugs and performance regressions. Reproduce → minimise → hypothesise → instrument → fix → regression-test. Use when user says "diagnose this" / "debug this", reports a bug, says something is broken/throwing/failing, or describes a performance regression.

Find deepening opportunities in a codebase, informed by the domain language in CONTEXT.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable.

Test-driven development with red-green-refactor loop. Use when user wants to build features or fix bugs using TDD, mentions "red-green-refactor", wants integration tests, or asks for test-first development.

Break a plan, spec, or PRD into independently-grabbable issues on the project issue tracker using tracer-bullet vertical slices. Use when user wants to convert a plan into issues, create implementation tickets, or break down work into issues.

Turn the current conversation context into a PRD and publish it to the project issue tracker. Use when user wants to create a PRD from the current context.

Sets up project rails for human + AI collaboration. Walks through five sections — CLAUDE.md/AGENTS.md coherence, issue tracker + triage labels, domain doc layout, code-style conventions, and version-control conventions — recording the result so the engineering skills (`to-issues`, `to-prd`, `diagnose`, `tdd`, `improve-codebase-architecture`, `zoom-out`) and any humans working alongside agents share the same baseline. Run on any new or existing repo before starting agent-driven work, or when those skills appear to be missing context.