Ask Matt

Ask Matt

You don't remember every skill, so ask.

A flow is a path through the skills. Most paths run along one main flow, and two on-ramps merge onto it. Everything else is standalone, or a vocabulary layer that runs underneath. Delivery tooling, including stacked PRs, sits underneath the main flow rather than replacing it.

The main flow: idea → ship

The route most work travels. You have an idea and want it built.

1. /grill-with-docs sharpens the idea by interview. Start here whenever you are working in a working directory: it's stateful, retaining what it learns in CONTEXT.md and ADRs. (No working directory? Use /grill-me instead, covered under Standalone. Both run the same /grilling primitive; grill-with-docs is the one that leaves a paper trail.)

2. Branch: can you settle every question in conversation? If a question needs a runnable answer (state, business logic, a UI you have to see), detour through a prototype, bridged by /handoff in both directions:

   • /handoff out, then open a fresh session against that file;
   • /prototype to answer the question with throwaway code;
   • /handoff back what you learned, and reference it from the original idea thread.

3. Branch: is this a multi-session build?

   • Yes → /to-spec (turn the thread into a spec), then /to-tickets to split it into tracer-bullet tickets, each declaring its blocking edges. On a local tracker that's one file per ticket under .scratch/<feature>/issues/, worked blockers-first by hand; on a real tracker the edges become native blocking links, so any ticket whose blockers are done can be grabbed: kick off /implement per ticket, clearing context between each one. Each ticket is self-contained, so the last one's context is disposable.
   • No → /implement right here, in the same context window.

   /implement is the user-invoked orchestrator. For non-trivial tickets it dispatches one isolated implementer subagent, which drives TDD at the agreed seams; the orchestrator then calls the Skill tool with "code-review" against the fixed point before delivery. Reach for /tdd on its own when you just want to build a concrete behaviour test-first without a full spec, and /code-review on its own whenever you want to review a branch or PR against a fixed point.

Delivery topology

After /to-tickets has produced and the user has approved the real blocker graph, choose the PR delivery topology:

• Linear dependency chain → invoke /implement in stacked mode and use the gh-stack skill for branch and PR operations.
• Fork or join with one final merge into trunk → invoke /implement in integration mode. Create one durable integration branch and target parallel ticket PRs at it; the final integration PR targets trunk and closes the spec/PRD.
• Fork or join without a final integration PR → ask whether to serialize the work into one linear stack or preserve parallelism with independent/dependent PRs. gh-stack supports only one parent and one child per stack.
• Existing PR or branch → adopt compatible existing state with gh stack init <branch> or use gh stack link only when branches are already managed by an external tool. Never recreate an existing PR just to introduce a stack.

The issue dependency graph remains the source of truth. A stack ordering is a delivery decision and must not silently rewrite a ticket's Blocked by edges.

When the user asks for parallel worktrees and a single final PRD merge, the integration branch is the default delivery topology. Child PRs target the integration branch and use Refs #<ticket>; only the final PR targets trunk and uses Closes #<spec>.

When using a stack, each fresh /implement context must run gh stack view --json and verify the current parent before coding. When using integration delivery, each fresh context must verify the integration branch and that all ticket blockers are merged into it. The branch state is durable; conversation context is not.

Context hygiene

Keep steps 1–3 in one unbroken context window (don't compact or clear until after /to-tickets) so the grilling, spec, and tickets all build on the same thinking. Each /implement then starts fresh, working from the ticket and the durable repository/issue state.

The limit on this is the smart zone: the window (~150k tokens on state-of-the-art models) within which the model still reasons sharply. If a session approaches it before /to-tickets, don't push on degraded: compact at the nearest phase boundary and carry on.

On-ramps

A starting situation that generates work, then merges onto the main flow.

• Bugs and requests piling up → /triage. It moves issues through triage roles and produces agent-ready issues, which /implement later picks up.

  Triage is only for issues you didn't create: bug reports, incoming feature requests, anything that arrives raw. Tickets that /to-tickets produced are already agent-ready, so don't triage them.

• Something's broken → /diagnosing-bugs. For the hard ones: the bug that resists a first glance, the intermittent flake, the regression that crept in between two known-good states. It refuses to theorise until it has a tight feedback loop (one command that already goes red on this bug), then fixes with a regression test.

• A huge, foggy effort: a greenfield project or a huge feature build, too big for one session → /wayfinder, the most cognitively demanding flow here. When the way from here to the destination isn't visible yet, it charts a shared map of decision tickets on the issue tracker and resolves them one at a time, producing decisions, not deliverables, until the fog is pushed back and the way is clear. Where /grill-with-docs sharpens an idea you can hold in one session, wayfinder is for the idea you can't, and it's slower and denser, so save it for exactly that, never a well-scoped feature.

  When the map clears, it hands off, it doesn't build: merge onto the main flow at /to-spec, which collapses the map's linked decisions into a buildable plan, then /to-tickets and /implement as usual. Looping the map straight into /implement skips that collapse and throws the linked detail away, so go straight to /implement only when the effort turned out genuinely small.

Codebase health

Not feature work, just upkeep.

• /improve-codebase-architecture runs whenever you have a spare moment to keep the codebase good for agents to operate in. It surfaces deepening opportunities; picking one generates an idea you can take into the main flow at /grill-with-docs. It's the survey that finds the candidates; /codebase-design is the bench you design the chosen one on.

Vocabulary underneath

Two model-invoked references that run beneath the other skills, each the single source of truth for its vocabulary. Reach for them directly when the words, not the process, are the problem; or let the skills above pull them in.

• /domain-modeling sharpens the project's domain language: challenge a fuzzy term, resolve an overloaded word ("account" doing three jobs), and record a hard-to-reverse decision as an ADR. It's the active discipline /grill-with-docs drives to keep CONTEXT.md a clean glossary.
• /codebase-design is the deep-module vocabulary (module, interface, depth, seam, adapter, leverage, locality) for designing a module's shape: a lot of behaviour behind a small interface at a clean seam. /tdd and /improve-codebase-architecture both speak it.

Phase boundaries

A phase is a chunk of work inside a session: the grilling, the implementation, the QA. At the boundary between two of them you have five options, and picking between them is the fuzziest decision in this whole map.

• Continue: stay put. Costs nothing, loses nothing.
• /clear: empty the window when nothing here matters to what's next.
• /handoff: write a portable markdown file for a new harness, directory, colleague, or side task.
• Subagent: send a tightly scoped task to its own window and get a report back.
• /compact: compress this context and seed a fresh session. It is the default at the bottom of the tree, not the first reach.

Read PHASE-BOUNDARIES.md for the ordered tree. Make the decision at the boundary; mid-phase, continue or split the remaining work into subagents.

Standalone

Off the main flow entirely.

• /grill-me is the same relentless interview as /grill-with-docs, but stateless: it saves nothing locally and builds no CONTEXT.md. Reach for it when you are not working in a working directory.
• /grilling is the interview primitive itself: rounds, the frontier, facts are the agent's job and decisions are yours. Reach for it directly only when you want the interview with no wrapper around it.
• /resolving-merge-conflicts works an in-progress merge or rebase conflict hunk by hunk, resolving by intent traced to each side's primary source.
• /prototype is a small, throwaway program that answers one design question. Its answer folds into the real code; keep the prototype as a primary source when the implementation issue points to it.
• /research delegates reading legwork to a background agent and leaves a cited Markdown file in the repo. Research feeds the main flow; it does not replace it.
• /to-questionnaire writes a questionnaire when the missing information is in someone else's head rather than yours or the codebase.
• /wizard covers steps only a human can perform, such as provisioning, credentials, or a one-off cutover.
• /wait-what re-pitches a message that did not land, using the project's vocabulary.
• /teach supports learning a concept over multiple sessions in the current directory.
• /writing-for-agents is the reference for writing documents agents consume: skills, AGENTS.md, and pointed-at docs.

Precondition

/setup-matt-pocock-skills runs before the first engineering flow to configure the issue tracker, triage labels, and doc layout the other skills assume. Custom issue trackers also work.