Browse Skills

Grill Me runs a /grilling session to pressure-test a plan or design with pointed questions. Use it when you want to surface gaps, assumptions, and weak spots before moving forward.

Grilling interviews you one question at a time to stress-test a plan or idea and reach a shared understanding. It looks up facts from the environment when possible and leaves decisions for you to confirm.

Implement carries out planned work from a spec, ticket, or agreed conversation into code. Use it when you want an existing plan turned into the smallest coherent change and verified with tests or checks.

Improve Codebase Architecture analyzes a codebase for high-leverage structural changes and explains the evidence, tradeoffs, and next steps. Use it for architecture audits or when deep-module refactoring may reduce friction.

Skill Router scans the active repository to identify stack and workflow signals, then recommends a capped set of skills, agents, and MCP servers. It also suggests unloads for helpers that no longer fit after confirmation.

Skill Router reads the current project's stack and loads or unloads skills based on context. Use it at session start or when switching projects so the active skill set matches the work.

Teach adapts explanations and exercises to the user’s current level and goal. It can also keep a lightweight learning record across sessions when that helps.

To Spec synthesizes an implementation-ready spec from a conversation and relevant repository context. Use it when work needs to be captured as a PRD or scoped specification with behavior, constraints, and acceptance criteria.

AC Autonomous Orchestrator coordinates autonomous coding operations from initialization through planning, execution, validation, and commit. Use it to run full-session implementations or manage multiple components together.

AC Handoff Creator packages current progress, decisions, blockers, and next actions so a later session can resume work. Use it when ending a session or saving resumable context.

Agent Factory helps define, build, and validate a specification-first digital FTE that composes earlier skills into a shippable agent. It is used when turning a customer problem into an execution skill with acceptance tests and production constraints.

Agent Integration helps connect custom LLM backends to agent frameworks using OpenAI-compatible APIs, tool calling, and MCP servers. Use it when configuring LiteLLM, SDK base URLs, or fallback paths for a local or hosted model.