.agents

Normalize a consequential or ambiguous request when missing context, boundaries, or proof could materially change the work. Explicit invocation previews the improved prompt.

Explore requirements and approaches through collaborative dialogue, then write a right-sized requirements document. Use when the user says "let's brainstorm", "what should we build", or "help me think through X", presents a vague or ambitious feature request, or seems unsure about scope or direction -- even without explicitly asking to brainstorm.

Find root causes and fix bugs. Use for errors, failing tests, issue repros, stack traces, "debug this", or "why is this failing".

Create structured plans for multi-step tasks once the goal is clear enough to plan. Use after ce-brainstorm or ce-grill, or directly for clear planning requests. If the request has branchy product/scope ambiguity, run ce-grill first. After markdown plans, document-review runs before handoff.

Review recent code changes for bugs, regressions, product fit, conventions, performance, security, and blast radius.

Execute implementation work with a compact product-contract loop. Use when the user asks to build, fix, implement, polish, or finish a scoped task. Favor reasoning, prior art, smallest correct changes, focused tests, and real-surface proof. Do not run autonomous PR, CI, ticket, or residual-work pipelines.

Route project knowledge searches to relevant collections on this machine and classify evidence authority. Use for non-trivial framing and planning or when implementation intent, ownership, conventions, expected behavior, or history are unclear; skip exact grounded work. Delegate query and retrieval mechanics to the qmd skill.

Refresh stale .ai/solutions learnings against current code. Use for "refresh my learnings", audit stale learnings, or consolidate overlapping docs.

Document a recently solved problem to compound your team's knowledge or CONCEPTS.md, the project's shared domain vocabulary.

Split or rebuild oversized diffs into reviewable pieces. Use for 'too big to review', 'break this down', 'decompose this', or 'split this up'.

Clarify and stress-test a requirements doc, plan, design, or rough approach through a focused interview. Use when the user wants to be grilled, or before ce-plan when the requirements are too branchy, ambiguous, or strategically important to plan cleanly.

Generate and evaluate grounded ideas. Use for 'what should I improve', 'give me ideas', 'ideate', 'surprise me', or 'what would you change'.

Improve CE skills from evidence. Use when the user asks to run a continual-improvement loop, reinforce, tune, patch, or evaluate ce-* skills using Codex/Claude/Cursor session history, QMD agent retros, gotchas, proposals, user corrections, retries, test failures, or recurring skill friction.

Search coding agent session history across Claude Code, Codex, and Cursor. Use for past work, previous attempts, or prior investigations.

Resolve a PR that conflicts with its base branch by rebasing onto the base and force-pushing with lease — never a merge commit. Use when a PR is out of date or has conflicts with main, when GitHub shows "This branch has conflicts that must be resolved", or when the user asks to rebase a branch, update a branch against main, or resolve PR conflicts.

Open Codiff for a narrative code walkthrough, a blocking plan handoff, or a shared walkthrough URL. Use when the user writes "$codiff", "/codiff", "$codiff plan", "$codiff share", "show me codiff", "open Codiff", or asks to review a change or edit a plan in Codiff.

Run any question, idea, or decision through a council of 5 AI advisors who independently analyze it, peer-review each other anonymously, and synthesize a final verdict. Use when the user asks for ce-council, council this, run the council, war room this, pressure-test this, stress-test this, debate this, validate this, get multiple perspectives, or when they are torn between meaningful options. Do not trigger on simple factual questions, trivial yes/no questions, implementation tasks, or casual should-I questions without a real tradeoff.

Run the changed-code quality gate after implementation: touched-file lint, format, type-check, tests, and code-shape checks. Use after ce-work writes code, before review or shipping, and when the user asks to make changed files clean.

Run a rigorous multi-pass technical code review of recent changes. Use when the user asks for ce-technical-review, a technical review, a deep code review, quadruple-checking implementation quality, or a strict pre-commit audit focused on correctness, performance, traceability, elegance, conventions, security, adversarial cases, and blast radius.

Recheck your thinking across the whole conversation. Works on any answer, idea, claim, or opinion worth a second pass.

Simplify and refine recently changed code for clarity, reuse, quality, and efficiency while preserving behavior.

Build with rpce-cli context builder plan → implement

Deep planning workflow using rpce-cli: map seams, draft, critique, polish — produces a ready-to-execute plan document

Deep investigation with rpce-cli commands: tools gather evidence, follow-up reasoning synthesizes selected context

Iterative performance optimization loop using rpce-cli: instrument with debug-only metrics, establish a baseline, then plan → delegate one optimize+harden cycle → re-measure → ask oracle for next plan, looping until the oracle is satisfied or the target metric is met

Export a ChatGPT-ready Question / Plan / Review prompt using rpce-cli

Plan, decompose, and delegate complex tasks across multiple agents using rpce-cli

Refactoring assistant using rpce-cli to analyze and improve code organization

Reminder to use rpce-cli

Code review workflow using rpce-cli git tool and context_builder

Apply TypeScript, React, and code-shape taste for implementation, review, refactor, and quality-gate tasks. Use when code changes involve TS/React conventions, typed contracts, helper boundaries, typed domain cores, IO or UI adapters, schemas, reducers, selectors, adapters, premature abstractions, speculative guards, maintainability smells, courier objects, JSX business logic, casts, useEffect, effect soup, or over-engineering.

Run metric-driven optimization loops with experiments and hard gates. Use for search, clustering, build, prompt, or measurable quality tuning.

Search Slack for interpreted org context and synthesize decisions, constraints, and discussion arcs.

Run an extremely strict maintainability review for abstraction quality, giant files, and spaghetti-condition growth. Use for a thermo-nuclear code quality review, thermonuclear review, deep code quality audit, or especially harsh maintainability review.

Define or refine a concrete measurable goal. Use for the goal tool, objectives, success criteria, or fuzzy intentions.

Find architecture deepening opportunities. Use to improve structure, testability, module boundaries, or agent navigability.

Audit a codebase read-only and produce prioritized implementation plans. Use for bugs, security, performance, tests, tech debt, DX, or roadmap ideas.

Use the Obsidian CLI for vault notes, tasks, properties, search, and plugin/theme debugging.

Generate or refresh ONBOARDING.md for a codebase. Use for onboarding docs, new contributor docs, 'vonboard', or 'vonboarding'.

Generate an interactive PR review walkthrough as an HTML page. Fetches PR data via gh API, categorizes files into core vs mechanical changes, adds reviewer annotations, and renders diffs with moved-code detection. Use when the user pastes a GitHub PR URL and asks for a review, walkthrough, or summary, or says "review this PR".