protocols-plugin

Multi-hypothesis research campaign protocol. Use for open-ended optimization goals ("make X faster", "improve the metric") where several candidate ideas compete for a limited budget. Triages ideas cheaply, confirms survivors with full preregistered experiments, and keeps a committed ledger so dead ideas stay dead.

Controlled benchmark experiment protocol for solver, performance, routing, and optimization work. Use when testing a hypothesis against metrics, comparing implementations, or deciding whether to promote or abandon a benchmark-driven change.

Coordination-only protocol. You dispatch subagents, you do not execute code. Restricts tools to Read, Write (prompts/ and notes- only), Agent, Glob, Grep. Use when orchestrating multi-agent work.

Four-role sequential pipeline for complex, high-risk changes. Scout surveys the codebase, Coder implements with TDD, Analyst finds problems, Verifier gates the merge. Use for large refactors, architectural shifts, and interdependent multi-phase work.

Background knowledge for writing and launching subagent prompts. Covers prompt file conventions, worker identity declarations, batch dispatch patterns, and parallel swarm safety. Auto-invocable when dispatching agents.

Deletion-first fixed-point protocol for refactors, cleanup, migrations, and architectural convergence. Use when replacing old surfaces, removing helper debt, shrinking code, moving ownership boundaries, or auditing every symbol in a slice.

Read-only design review against project principles. Checks whether a design or implementation aligns with or violates the project's stated principles. Does not check code quality, bugs, tests, style, or performance — only directional alignment.

Use Codex and Gemini CLIs as external reviewers for gated review chunks. They read prompt files and write reports. Use for pre-implementation review, spec validation, architecture critique, or second opinions.

Structured debugging with competing hypotheses and escalation levels. Use when encountering unexpected behavior, multiple possible causes, or after 3+ failed fix attempts. Separates facts from theories and tracks evidence systematically.

Tracked iteration cycles for reducing failures. Number iterations sequentially, record state before/after, commit atomically per fix. Use when iteratively reducing test failures, fixing bugs in cycles, or any work requiring tracked progress across multiple attempts.

Orchestrate multi-phase workflows with parallel and sequential stages. Research, plan, implement, verify — each phase produces filesystem artifacts consumed by the next. Use when task decomposes into stages that would overflow a single agent's context.

Reverse engineering protocol. The work product is documentation. Decompile function, document findings in NOTES.md, then next function. Never batch discoveries, never mark phase complete without reviewing NOTES.md against phase requirements.

Pre-implementation research protocol. Gather information, explore unfamiliar codebases/APIs, answer "how does X work" questions. Has web access. Creates structured findings in docs/reports/ for later synthesis. Use before any non-trivial implementation.

Spec update workflow with discovery, drafting, external review gates (Codex + Gemini), and integration phases. Use when adding, modifying, or extending specification documents. Specs are truth — code implements specs, never the reverse.