skills

Scaffold or repair `etc/http/` OpenCollection trees for Bruno. Use for new collections, brownfield standardization, or Bruno/Postman migration.

Control herdr (a terminal-native agent multiplexer) from inside it. Manage workspaces and tabs, split panes, spawn sibling agents, read pane output, and wait for state changes — all via CLI commands that talk to the running herdr instance over a local unix socket. Use when running inside herdr (HERDR_ENV=1). Do not use outside herdr.

Sets up an `## Agent skills` block in AGENTS.md/CLAUDE.md and `docs/agents/` so the engineering skills know this repo's issue tracker (GitHub, GitLab, fp, or local markdown), triage label vocabulary, and domain doc layout. Run before first use of `fp-plan`, `fp-implement`, `fp-review`, `to-issues`, `to-prd`, `triage`, `diagnose`, `tdd`, `improve-codebase-architecture`, or `zoom-out` — or if those skills appear to be missing context about the issue tracker, triage labels, or domain docs.

Grilling session that challenges your plan against the existing domain model, sharpens terminology, and updates documentation (CONTEXT.md, ADRs) inline as decisions crystallise. Use when user wants to stress-test a plan against their project's language and documented decisions.

Find deepening opportunities in a codebase, informed by the domain language in CONTEXT.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable.

Build a throwaway prototype to flush out a design before committing to it. Routes between two branches — a runnable terminal app for state/business-logic questions, or several radically different UI variations toggleable from one route. Use when the user wants to prototype, sanity-check a data model or state machine, mock up a UI, explore design options, or says "prototype this", "let me play with it", "try a few designs".

Stress-test a plan against the domain model. Interrogates ambiguous terms, verifies claims against code, and updates CONTEXT.md, CONTEXT-MAP.md, and docs/adr/ inline. Do NOT use for greenfield ideation, code review, or AGENTS.md.

Deep codebase exploration across remote repositories. Fetch, search, and compare library source code. Use for internals, architecture analysis, and cross-repo patterns. Skip simple API or docs questions.

Fetch dependency source code to give AI agents deeper implementation context. Use for library internals, dependency debugging, and real implementation review. Skip simple API docs; route multi-repo work to librarian or context7.

Resolves TypeScript type errors, eliminates `any`, and designs advanced generic types. Use when asked about TS types, generics, `infer`, conditional types, mapped types, template literal types, branded types, type narrowing, utility types, or diagnosing type errors. Do not use for runtime JS logic, Node.js APIs, bundler config, or ESLint rules.

Find, claim, and complete work on fp issues with session handoff. Activate when the user says "work on issue", "pick up task", "what should I work on", "use fp", or "tackle next task". Not for planning or reviewing.

Create plans and break them down into trackable fp issue hierarchies with dependencies. Activate when the user says "create a plan", "break down feature", "use fp", or "structure this work". Not for implementing tasks or reviewing code.

Review code changes, assign commits to issues, leave structured feedback, and create stories. Activate when the user says "review code", "assign commits", "create a story", "use fp", or "prepare for review". Not for planning or implementing tasks.

Triage GitHub issues through a label-based state machine with interactive grilling sessions. Activate when the user says "triage issues", "review incoming issues", or "what needs triage". Not for creating issues, writing code, or PR reviews.

Measure and fix AI skill activation, clarity, and cross-agent reliability. Use when skill uptake is weak, scores regress, or context is bloated. Not for writing new skills from scratch or general prompt engineering.

Extract a validated problem brief through user interview and codebase exploration. Activate ONLY when the user says "new PRD", "plan a feature", or "write a PRD". Not for implementation specs, bug triage, or general feature discussion.

Continuation skill: assemble a PRD from validated artifacts, run quality gates, and publish. Runs after prd-interview when both problem-brief.md and decision-ledger.md exist. Not for writing PRDs from scratch.

Continuation skill: walk the design tree branch-by-branch until every decision is explicit. Runs after prd-discover when problem-brief.md exists. Not for standalone use, brainstorming, or solution design.

Break a PRD into dependency-ordered vertical slice GitHub issues. Activate when the user says "break down PRD", "create issues from PRD", or "slice the PRD". Not for writing PRDs, triaging bugs, or creating non-implementation issues.

Produce a DDD Context Map showing relationships between Bounded Contexts from existing UBIQUITOUS_LANGUAGE.md glossaries or conversation context. Saves to CONTEXT_MAP.md. Triggers on "context map", "context mapping", "bounded context relationships", "how do these contexts relate", "upstream downstream".

Index a codebase for agent code understanding using codemogger (semantic + keyword search) and cx (symbol navigation). Optionally add zoekt (trigram regex). Use when preparing for architecture analysis, exploring an unfamiliar codebase, onboarding to a project, or when grep/ripgrep returns too many irrelevant matches. Do NOT use if indexes already exist and are current, for non-code assets, or for single-file lookups where grep suffices.

Apply Feature-Sliced Design (FSD) architecture when creating or modifying frontend code. Routes code to correct layers (app/pages/widgets/features/entities/shared) and enforces strict downward-only import rules. Use when creating components, pages, features, or restructuring React/Next.js/Vue/Remix projects. Do NOT use for backend services, non-UI libraries, single-file scripts, or projects with fewer than 5 components.

Design multiple radically different interfaces for a module-deepening candidate and create a refactor RFC as a GitHub issue. Use after explore-architecture surfaces candidates and the user picks one. Do NOT use without a specific candidate chosen, for simple extract-function refactors, or without prior coupling evidence from explore-architecture.

Search an indexed codebase to find architectural friction and surface module-deepening candidates. Use after index-codebase has run, when the user asks to find refactoring opportunities, identify shallow modules, assess coupling, or improve testability. Do NOT use for greenfield architecture design, performance optimization, or before the codebase is indexed with codemogger.

Build features or fix bugs using strict red-green-refactor TDD with vertical slices. Use when the user wants test-first development, mentions "red-green-refactor", wants integration-style tests, or asks to build something with TDD. Do NOT use for writing tests after implementation, adding tests to untested legacy code retroactively, or pure unit testing of isolated functions.

Investigate a bug by searching the codebase for root cause, then create a GitHub issue with a TDD fix plan. Use when the user reports a bug, mentions "triage", wants to investigate a problem, or needs a fix plan filed as an issue. Do NOT use for feature requests, architecture improvements, problems without observable symptoms, or when the user already knows the root cause and just wants to fix it.

Structured dialectical reasoning engine — two subagents believe opposed positions at full conviction (Electric Monks) while the orchestrator performs structural contradiction analysis and synthesis (Aufhebung). Use when stress-testing an idea, resolving genuine tension, making high-stakes decisions where tradeoffs are unclear, or building a deeper mental model of a domain. Works across technical architecture, product strategy, philosophy, personal decisions, risk analysis, and policy.

Proactively suggest diagrams when explaining complex systems. Triggers on diagrams, charts, visualizations, flowcharts, sequence diagrams, architecture diagrams, ER diagrams, state machines, Gantt charts, mindmaps, C4, class diagrams, git graphs. Use when user asks for visual representations of code, systems, processes, data structures, database schemas, workflows, or API flows. Generate Mermaid diagrams in markdown.

Creates new AI agent skills following the Agent Skills spec. Load FIRST before writing any SKILL.md. Provides required format, naming conventions, progressive disclosure patterns, and validation. Use when building, reviewing, or debugging skills.

Set up Claude Code hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations, add git safety hooks, or block git push/reset in Claude Code.

Use this skill when user wants to create a refactor plan.

Console.log debugging patterns for React/TypeScript. Use when tracking down bugs involving data transformations, conditional logic, state management, or when you need to trace data flow through pipelines.

Jujutsu (jj) version control commands and workflows. Use when performing version control operations, committing changes, managing branches/bookmarks, or syncing with remotes. Always use jj instead of git.

Dialogue-driven spec development through skeptical questioning and iterative refinement. Triggers: 'spec this out', feature planning, architecture decisions, 'is this worth it?' questions, RFC/design doc creation, work scoping. Invoke Librarian for unfamiliar tech/frameworks/APIs.

Search the public web using the Exa-backed websearch tool. Use this when the user asks for up-to-date facts, external docs, release notes, APIs, or anything not reliably present in the local repo.