eng

Cancel active Ship Loop

Help engineers prepare decision-focused briefs (DECIDE.md) for 1:1 meetings with technical leadership. Conversational extraction of decisions, options, uncertainty, and supporting evidence into a structured brief. Use when preparing for a 1:1, decision brief, DECIDE.md, or when an engineer needs help articulating what they need from leadership. Triggers: 1on1, 1:1, prep 1:1, prepare for 1:1, 1on1 prep, decision brief, DECIDE.md, prepare for meeting with CEO/CTO.

Systematic debugging for local development. Enforces root cause investigation — never implements fixes. 5-phase: Triage, Reproduce & Comprehend, Investigate, Classify, Report & Recommend. Classifies root causes as dev environment/config vs code bugs. Uses tools aggressively for observable verification — runs commands, queries state, checks actual output. Includes triage playbooks, hypothesis-test-refine cycles, tool patterns, agent metacognition (loop detection, strategy switching, confidence calibration), escalation heuristics. Presents root cause findings with recommended fix strategy and hands off to implementer. Two autonomy modes: Supervised (propose diagnostic mutations, wait for approval) and Headless (iterate freely within approved action classes). Composable with /implement, /qa, /ship, /tdd, /explore. Triggers: debug, fix bug, root cause, why is this failing, investigate error, diagnose, troubleshoot, something broken, test failure, crash, regression, stack trace, error message, it worked before, flaky test, wrong output, not working, build failure, type error, exception, debugging.

Decompose a SPEC.md into spec.json (structured user stories with dependency ordering and verifiable acceptance criteria). Use when breaking down a feature spec into implementable tasks, reviewing a decomposition before implementation, or re-decomposing a story that proved too complex. Triggers: decompose, task breakdown, spec.json, split stories, break down spec.

Write or update documentation for engineering changes — both product-facing (user docs, API reference, guides) and internal (architecture docs, runbooks, inline code docs). Builds a world model of what changed and traces transitive documentation consequences across all affected surfaces. Discovers and uses repo-specific documentation skills, style guides, and conventions. Standalone or composable with /ship. Triggers: docs, documentation, write docs, update docs, document the changes, product docs, internal docs, changelog, migration guide.

Build deep, structured understanding of a codebase area or system topic before acting on it. Three complementary lenses: (1) Surface mapping — enumerate product and internal surfaces a topic touches, trace connections and blast radius across the system. (2) Pattern inspection — discover conventions, shared abstractions, and the "grain" of a code area by reading siblings. (3) System tracing — follow call chains, map transitive dependencies, identify cross-boundary contracts and implicit coupling. Use before spec work, implementation, review, debugging, or any task that needs understanding of what a feature touches and how the code works. Subsumes find-similar. Triggers: explore, inspect, discover, what does this touch, map the surfaces, world model, surface areas, what's affected, what patterns exist here, how does this area do X, before implementing, check conventions, find similar, trace the flow, blast radius, what would break, dependency tree, codebase inspection, feature scope, what would this change.

Convert SPEC.md to spec.json, craft the implementation prompt, and execute the iteration loop via subprocess. Use when converting specs to spec.json, preparing implementation artifacts, running the iteration loop, or implementing features autonomously. Triggers: implement, spec.json, convert spec, implementation prompt, execute implementation, run implementation.

Spawn child Claude Code instances from a parent Claude Code session. Covers the env guard pattern, stdin redirection, parallel execution, monitoring, and result collection. Use when you need to run multiple independent Claude Code processes concurrently for tasks like parallel research, bulk code review, multi-repo operations, or any work that benefits from full isolated Claude Code sessions rather than lightweight Task-tool subagents. Triggers: spawn claude, nested claude, parallel claude, child process, subprocess claude, run N instances, concurrent agents, fan-out.

Write or update a GitHub PR body — self-contained, stateless, and structured for efficient review. Gathers context from the diff, spec, and implementation to produce a comprehensive PR description. Standalone or composable with /ship and /review-cloud. Triggers: PR body, PR description, pull request, write PR, update PR, create PR, PR template.

Take SPEC.md + code + diff and produce a maximally ambitious, user-outcome-grounded qa-progress.json ready for /qa execution. Mock-aware coverage reality check — mocked tests don't count as real coverage. Derives scenarios directly from SPEC.md source material (user journeys, failure modes, acceptance criteria), synthesizes compositional user session scenarios, traces spec-to-code and code-to-spec, investigates code patterns, prioritizes, and writes the plan. No execution boundaries — plans everything worth testing including browser, Docker, and integration scenarios. Standalone or composable with /ship. Triggers: qa-plan, qa plan, test plan, derive qa scenarios, qa planning, plan qa, test scenarios.

Manual QA testing — verify features end-to-end as a user would, by all means necessary. Exhausts every local tool: browser (Playwright), Docker, ad-hoc scripts, REPL, dev servers. Mock-aware — mocked test coverage does not count. Proves real userOutcome at highest achievable fidelity. Blocked scenarios flow to /pr as pending human verification. Standalone or composable with /ship. Triggers: qa, qa test, manual test, test the feature, verify it works, exploratory testing, smoke test, end-to-end verification.

Manage the PR review iteration loop on GitHub: poll for reviewer feedback, assess suggestions with evidence, implement fixes, resolve threads, and drive CI/CD to green. Use after pushing a branch and creating a PR. Works standalone on any PR (bug fix, refactor, hotfix, feature). Spec-aware when a SPEC.md is provided, but does not require one. Triggers: review-cloud, review cloud, PR feedback, iterate on PR, resolve reviews, CI/CD resolution, green pipeline, address reviewer comments.

Run the local AI review convergence loop on the current branch. Dispatches 17 domain-specific reviewers, produces a structured review summary, and runs a bounded fix convergence loop when findings are blocking. Use for AI code review before pushing, as a quality checkpoint mid-development, or standalone outside /ship. Triggers: review-local, local review, local code review, AI review, pre-push review, review my code locally.

Orchestrate any code change from requirements to review-ready branch — scope-calibrated from small fixes to full features. Composes /spec, /implement, and /research with depth that scales to the task: lightweight spec and direct implementation for bug fixes and config changes, full rigor for features. Produces tested, locally reviewed, documented code on a feature branch. The developer pushes the branch and creates the PR. Use for ALL implementation work regardless of perceived scope — the workflow adapts depth, never skips phases. Triggers: ship, ship it, feature development, implement end to end, spec to PR, implement this, fix this, let's implement, let's go with that, build this, make the change, full stack implementation, autonomous development.

Drive an evidence-driven, iterative product+engineering spec process that produces a full PRD + technical spec (often as SPEC.md). Use when scoping a feature or product surface area end-to-end; defining requirements; researching external/internal prior art; mapping current system behavior; comparing design options; making 1-way-door decisions; negotiating scope; and maintaining a live Decision Log + Open Questions backlog. Triggers: spec, PRD, proposal, technical spec, RFC, scope this, design doc, end-to-end requirements, scope plan, tradeoffs, open questions.

Universal test-driven development methodology and test design craft. Background knowledge for writing behavior-focused tests, mocking strategy, vertical slicing, and tracer bullet execution. Auto-loaded when writing tests during implementation, review feedback, or standalone test authoring. Triggers: writing tests, test design, mocking strategy, test quality, behavior testing, integration testing, test-first, red-green-refactor.

VP of Product + VP of Engineering reviewer. Steers AI-generated specs, research, analysis, and designs through Socratic questioning, evidence demands, and principle-based challenges. Acts as a senior technical/product leader who elevates and audits knowledge work. Use when reviewing spec output, research reports, analysis findings, architecture proposals, or any structured knowledge work that needs senior leadership-caliber scrutiny.

4 hooks across 3 events