laboratory

Assess and improve codebase readiness for AI coding agents. Dispatches parallel subagents to evaluate style, testing, docs, architecture, CI, observability, security, and dev environment. Produces a scored report with prioritized remediation. Then executes the highest-impact fixes. Use when: "agent readiness", "is this codebase agent-ready", "readiness report", "make this codebase agent-friendly", "agent-ready assessment", "readiness audit", "prepare for agents". Trigger: /agent-readiness, /readiness.

Dialectical premise-and-alternatives audit for a plan, spec, or context packet. Four moves: premise challenge (is this the right problem?), mandatory structurally-distinct alternatives, cross-model outside voice, user-ratified convergence. Operationalizes the AGENTS.md "Diverge Before You Converge" doctrine at the plan-review stage. Use when: about to commit to a plan/spec/design, reviewing a ticket before shape, when "is this the right problem" would be useful, or any time a proposal smells like a symptom fix instead of a root-cause fix. Trigger: /ceo-review, /challenge, /premise-check.

Laboratory CI gate audit and verification workflow. Use when checking whether experiment, harness, analysis, docs, or Spellbook harness changes are safe to ship. Trigger: /ci, /gates.

Parallel multi-agent review for Laboratory changes. Prioritizes correctness, scientific validity, provenance, tests, and harness fit. Trigger: /code-review, /review, /critique.

Take one Laboratory backlog item to merge-ready. Composes /shape, /implement, /code-review, /ci, /refactor, and /qa as needed. Stops before merge. Trigger: /deliver.

Capture and share what changed — every app has a demo path; the format depends on the change and the audience. The spectrum runs from a release-notes blurb ("no demo needed, here's what moved") through a curl + JSON paste, a terminal recording, a screenshot, a GIF, up to a polished narrated launch video. Pick the path that fits the change shape (UI / API / CLI / library / agent tool / internal-only), the audience (PR reviewer / team Slack / external customer), and the time budget. Also scaffolds a project-local demo skill for repos with recurring polished-capture needs. Use when: "make a demo", "generate demo", "record walkthrough", "launch video", "PR evidence", "upload screenshots", "demo artifacts", "make a video", "demo this feature", "create a walkthrough", "show what changed", "paste the output", "release notes blurb", "scaffold demo", "generate demo skill". Trigger: /demo.

Investigate, audit, triage, and fix. Systematic debugging, incident lifecycle, domain auditing, and issue logging. Four-phase protocol: root cause → pattern analysis → hypothesis test → fix. Use for: any bug, test failure, production incident, error spikes, audit, triage, postmortem, "diagnose", "why is this broken", "debug this", "production down", "is production ok", "audit stripe", "log issues". Trigger: /diagnose.

Outer loop for Laboratory research work. Picks a backlog item, shapes it, delivers it, settles it, ships it through PR, monitors evidence, and reflects. Trigger: /flywheel.

Always-on backlog grooming. Tidy, brainstorm, interrogate, investigate, research, and simplify in a single loop. Tidy is not a mode — it happens every time. Strategic-layer work fans out parallel subagents: /office-hours on raw items, /ceo-review on shaped packets, a technical-review bench on code hotspots, /research for external context. Use when: "groom", "what should we build", "rethink this", "biggest opportunity", "backlog", "prioritize", "backlog session". Trigger: /groom, /backlog, /rethink, /moonshot, /scaffold.

Build, maintain, evaluate, and optimize the agent harness — skills, agents, hooks, CLAUDE.md, AGENTS.md, and enforcement infrastructure. Use when: "create a skill", "update skill", "improve the harness", "sync skills", "eval skill", "lint skill", "tune the harness", "add skill", "remove skill", "convert agent to skill", "audit skills", "skill health", "unused skills", "lint this skill", "eval this skill", "evaluate this skill", "validate this skill", "check this skill". Trigger: /harness, /focus, /skill, /primitive.

Atomic TDD build skill. Takes a context packet (shaped ticket) and produces code + tests on a feature branch. Red → Green → Refactor. Does not shape, review, QA, or ship — single concern: spec to green tests. Use when: "implement this spec", "build this", "TDD this", "code this up", "write the code for this ticket", after /shape has produced a context packet. Trigger: /implement, /build (alias).

Four behavioral guidelines to reduce common LLM coding mistakes: surface assumptions, prefer simplicity, make surgical changes, drive by verifiable goals. Distilled from Andrej Karpathy's observations on where LLM agents silently fail. Reference inline when facing a judgment call about scope, simplicity, assumptions, or success criteria. Use when: "am I overcomplicating this", "should I refactor this adjacent code", "what are my assumptions", "how do I verify success", "/karpathy", "/principles".

Research, compare, and select LLM models for AI-powered apps and workflows. Finds latest models per family, verifies availability on target platform, compares pricing/benchmarks/tool-calling, produces ranked recommendations. Use when: "which model", "compare models", "find a model", "model research", "best model for", "cheapest model", "tool calling models", "model selection", "upgrade model", "swap model", "fallback models", "model chain".

Watch Laboratory signals after a run, analysis change, publication update, CI change, or repeated agent workflow. Uses local files, CI status, data integrity checks, and experiment artifacts rather than deploy telemetry. Trigger: /monitor.

YC-partner-style interrogation of a raw idea. Six forcing questions before you shape anything: demand reality, status quo, desperate specificity, narrowest wedge, observation & surprise, future-fit. Operationalizes the AGENTS.md "Diverge Before You Converge" doctrine at the ideation stage — problem-diamond, pre-/shape. Use when: user arrives with a rough idea, backlog item is fuzzy, you can't already write a one-sentence goal with a testable outcome, or the phrase "what should we build" would be useful. Trigger: /office-hours, /oh, /interrogate.

Verify laboratory experiment changes beyond "tests passed": simulate runs, analysis outputs, trace/provenance checks, and publication-readiness evidence. Use when: "QA this experiment", "verify this harness", "check publishability", "capture evidence", "smoke test this analysis". Trigger: /qa.

Branch-aware simplification and refactoring workflow. On feature branches, compare against base and simplify the diff before merge. On primary branch, scan the full codebase, research prior art, and identify the highest-impact simplification opportunity. Use when: "refactor this", "simplify this diff", "clean this up", "reduce complexity", "pay down tech debt", "make this easier to maintain", "make this more elegant", "reduce the number of states", "clarify naming". Trigger: /refactor.

Session retrospective, operator coaching, harness postmortem, codification, and outer-loop cycle critique. Distills learnings into hooks/rules/skills, mutates the backlog from evidence, and emits harness-tuning suggestions to a branch. Learning engine of the outer loop. Use when: "done", "wrap up", "what did we learn", "retro", "reflect", "calibrate", "how could I have asked better", "prompt better", "teach me from this session", "what should I learn from this", "reflect on cycle", "cycle postmortem", post-/flywheel critique. Trigger: /reflect, /retro, /calibrate, /reflect cycle <cycle-ulid>.

Web research, multi-AI delegation, and multi-perspective validation. /research [query], /research delegate [task], /research thinktank [topic]. Triggers: "search for", "look up", "research", "delegate", "get perspectives", "web search", "find out", "investigate", "introspect", "session analysis", "check readwise", "saved articles", "reading list", "highlights", "what are people saying", "X search", "social sentiment", "trending".

Polish a Laboratory branch until it is merge-ready. Runs code review, make ci-smoke, refactor, and experiment QA. Stops before merge. Trigger: /settle, /pr-fix, /pr-polish.

Shape a raw idea into something buildable. Product + technical exploration. Spec, design, critique, plan. Output is a context packet. Use when: "shape this", "write a spec", "design this feature", "plan this", "spec out", "context packet", "technical design". Trigger: /shape, /spec, /plan, /cp.

Final mile for a merge-ready Laboratory branch: verify landability, preserve backlog closure, merge through PR or git-native flow, update docs, reflect, and monitor evidence. Trigger: /ship.

Turn Laboratory worktree changes into coherent commits and push when appropriate. Classifies every path, keeps experiment/data boundaries clear, and refuses secrets or default-branch pushes. Trigger: /yeet, /ship-local.