kibertoad

Adds zone-based layout regions to the shell that modules populate per-route or per-active-tab. Use when the shell layout has dynamic panels, sidebars, or header regions that change based on the current route or active module.

Creates a per-entity scoped Zustand store using createScopedStore(). Use when you need independent state instances keyed by ID — per-interaction tabs, per-conversation messages, per-workspace scratchpads, etc.

Adds slot-based extensibility to the application — defines slot types in app-shared, contributes items from modules, and consumes them in the shell. Use when the shell needs to collect and render contributions from multiple modules (command palettes, system registrations, dashboard widgets, etc.).

Writes tests for a Reactive module using @tanstack-react-modules/testing with renderModule(), resolveModule(), and createMockStore(). Use when adding or updating tests for module components, routes, or slot contributions.

Adds a new shared dependency (zustand store, plain service, or reactive service) to the AppDependencies contract and wires it into the shell. Use when adding new cross-cutting concerns like analytics, feature flags, call adapters, or notification services.

Adds data fetching to a module component using React Query with Lokalise HTTP client and API contracts. Use when a component needs to fetch, create, update, or delete server data, or when integrating a new backend contract.

Creates a new Reactive module with routes, navigation, page components, and package.json. Use when asked to add a new feature module, section, or domain area to the application.

Registers an existing module in the shell host app. Use when wiring a new or existing module into the running application.

Adds slot-based extensibility to the application — defines slot types in app-shared, contributes items from modules, and consumes them in the shell. Use when the shell needs to collect and render contributions from multiple modules (command palettes, system registrations, dashboard widgets, etc.).

Adds data fetching to a module component using React Query with Lokalise HTTP client and API contracts. Use when a component needs to fetch, create, update, or delete server data, or when integrating a new backend contract.

Creates a per-entity scoped Zustand store using createScopedStore(). Use when you need independent state instances keyed by ID — per-interaction tabs, per-conversation messages, per-workspace scratchpads, etc.

Adds zone-based layout regions to the shell that modules populate per-route or per-active-tab. Use when the shell layout has dynamic panels, sidebars, or header regions that change based on the current route or active module.

Creates a new Reactive module with routes, navigation, page components, and package.json. Use when asked to add a new feature module, section, or domain area to the application.

Registers an existing module in the shell host app. Use when wiring a new or existing module into the running application.

Writes tests for a Reactive module using @react-router-modules/testing with renderModule(), resolveModule(), and createMockStore(). Use when adding or updating tests for module components, routes, or slot contributions.

Adds a new shared dependency (zustand store, plain service, or reactive service) to the AppDependencies contract and wires it into the shell. Use when adding new cross-cutting concerns like analytics, feature flags, call adapters, or notification services.

Run a cat-factory Pi-agent smoketest end to end. Use when asked to smoketest, sanity-check, or shake out a model on real coding tasks through the actual Pi setup — e.g. "smoketest the new Cloudflare model", "does Llama get stuck on the healthcheck task", "shake out qwen-coder on the coding fixtures". Configures the matrix, runs cat-smoke against actual Cloudflare AI, and surfaces breakage / dead-ends / loops from the captured transcripts. It does NOT grade quality (that's the benchmark skill).

Investigate a failed or suspicious agent run (implementer/coder, bootstrap, reviewer, etc.) using the production telemetry in D1. Use when asked to look into a run failure, "no progress" abort, a stuck/looping agent, model output quality, token/truncation issues, or "what went wrong with the latest run". Pulls the run lifecycle from `agent_runs` and the per-LLM-call prompt/response/usage from `llm_call_metrics`, then reads the tool-call loop to find the root cause.

Run a cat-factory agent benchmark end to end. Use when asked to benchmark, compare, or measure models or prompt versions on the agent tasks (requirement review, code review, implementation) — e.g. "benchmark Llama vs Claude on code review", "compare review@v1 vs a new prompt", "run the benchmark and grade it". Configures the matrix, runs cat-bench, invokes the benchmark-arbiter skill to grade, merges the report, and offers to commit under docs/benchmarks/.

Grade cat-factory benchmark runs. Use when asked to grade, score, judge, or arbitrate a benchmark run produced by the benchmark-harness (cat-bench) — typically a directory under docs/benchmarks/ containing a grading/ folder. Reads each cell's candidate output against its rubric and writes grades.json, grading-summary.md, and conclusions.md.

Boot cat-factory locally in no-auth dev mode, drive the running UI with Playwright to capture fresh screenshots, and embed them into the documentation pages that benefit. Use when asked to launch/spin up the app locally, refresh the product screenshots, or add UI images to the docs.

Synchronize this documentation site with the cat-factory code repo. Reads recent commit history from the local cat-factory checkout, finds product changes that affect deployers, operators, and agent/gate authors, updates the guide/deploy/reference docs to describe the current state, and opens a PR. Use when asked to sync docs, update docs from the code, or document recent changes.

Review a solution for bugs, implementation gaps, performance concerns, weak typing, low-value or redundant tests, and other code quality issues. Produces a prioritized list of findings with a recommendation on what to address.