Read this skill when working on the planner CLI's eval harness (apps/cli/evals) - the SDK-integration eval, the planner-step evals, the artifact bootstrap - or when adding/capturing a new eval case. Covers the two eval layers, how a case is structured, how to run each eval, and how to author a new case + rubrics.
Always read this skill before editing any file under packages/ai/ or packages/visual-ai/. Covers AI primitives: the agent abstraction, model registry, structured output generation, visual checkers, and element detection used across the platform.
Watch a PR's OpenCode review and notify the user when it finishes. Polls the OpenCode workflow run (opencode-review.yml, or the /oc-triggered opencode-comment.yml) until it settles - following supersede when a new push cancels an in-flight review - then reads OpenCode's comment, summarizes its findings, and pings the user via their preferred method (default: a completion chime + treemux attention + an in-chat summary). Use when the user wants to be told when the OpenCode bot is done reviewing a PR, e.g. 'watch the opencode review and let me know when it's done', 'ping me when opencode finishes on PR 1485'.
Always read this skill before editing any file under apps/ui/. Contains all UI conventions, data fetching patterns, design system rules, and component guidelines for the Vite + React 19 SPA.
Orientation and playbook for investigating a diff job (DiffsJob) execution from production logs and the database - the trigger chain, Temporal workflow lifecycle, data model, where logs and agent conversations live, and how to correlate them. Use when investigating a specific diff job or snapshot, a missing/stuck/duplicate snapshot, affected tests or newly authored tests, a generation/replay/refinement outcome, or any diffs-analysis / worker-diffs / refinement-loop behavior. Read REFERENCE.md for table schemas, query recipes, and Sentry recipes.
Always read this skill before editing any file under packages/ai/. Covers AI primitives: the agent abstraction, model registry, structured output generation, visual checkers, and element detection used across the platform.
Always read this skill before editing any file under packages/engine/. Covers the platform-agnostic execution agent, command system, driver interfaces, and runner that engine-web (Playwright) and engine-mobile (Appium) extend.
Read this skill when working on scenario provisioning, the @autonoma/scenario package, the customer SDK endpoint contract, scenario recipe ingestion, scenario instance lifecycle, or any code that reads/writes the Scenario / ScenarioInstance / ScenarioRecipeVersion / WebhookCall tables.