amplihack

Auto-generates code flow diagrams from multi-language module analysis. Detects when architecture diagrams become stale (code changed, diagram didn't). Supports Python, TypeScript/JavaScript, Rust, and Go out of the box. Use when: creating new modules, reviewing PRs for architecture impact, or checking diagram freshness across polyglot repositories. Generates mermaid diagrams showing imports, dependencies, and module relationships.

Execute multiple independent development tasks in parallel using subprocess isolation. Each task runs in a clean /tmp clone with Recipe Runner code-enforced workflow execution. Proven pattern: 4/5 PRs created successfully in first production use.

Interactive teaching agent for the goal-seeking agent generator and eval system. Provides a structured 14-lesson curriculum covering agent generation, SDK selection, multi-agent architecture, progressive evaluation (L1-L12), retrieval strategies, intent classification, math code generation, self-improvement loops with patch proposer and reviewer voting, and memory export/import.

Development workflow for features, bugs, refactoring. Normally executed as a sub-recipe by dev-orchestrator/smart-orchestrator. Supports direct invocation via recipe runner for standalone use.

Default task orchestrator for all development and investigation work. Classifies tasks, decomposes into parallel workstreams if appropriate, and routes execution through the recipe runner. Replaces ultrathink-orchestrator.

Writing clear, discoverable software documentation following the Eight Rules and Diataxis framework. Use when creating README files, API docs, tutorials, how-to guides, or any project documentation. Automatically enforces docs/ location, linking requirements, and runnable examples.

Generates comprehensive end-to-end test scenarios using outside-in methodology. Supports 5 app types: Web (Playwright), CLI, TUI, API, and MCP (gadugi YAML). Auto-detects app type or accepts explicit override.

Guides architects on when and how to use goal-seeking agents as a design pattern. This skill helps evaluate whether autonomous agents are appropriate for a given problem, how to structure their objectives, integrate with goal_agent_generator, and reference real amplihack examples like AKS SRE automation, CI diagnostics, pre-commit workflows, and fix-agent pattern matching.

6-phase investigation workflow for understanding existing systems. Normally executed as a sub-recipe by dev-orchestrator/smart-orchestrator. Supports direct invocation via recipe runner for standalone use.

Extracts key learnings from conversations, debugging sessions, and failed attempts. Use at session end or after solving complex problems to capture insights. Stores discoveries in memory (via amplihack.memory.discoveries), suggests PATTERNS.md updates, and recommends new agent creation. Ensures knowledge persists across sessions via Kuzu memory backend.

Automated Python-to-Rust migration via iterative convergence loops. Treats the Python codebase as the living specification and produces a fully-tested Rust equivalent with zero-tolerance parity validation. Use when migrating Python modules, libraries, or CLIs to Rust. Activates for: migration, oxidize, python to rust, port to rust, rewrite in rust.

Expert project manager orchestrating backlog-curator, work-delegator, workstream-coordinator, and roadmap-strategist sub-skills. Coordinates complex software projects through delegation and strategic oversight. Activates when managing projects, coordinating work, or tracking overall progress.

Iterative codebase quality audit with multi-agent validation and escalating-depth SEEK/VALIDATE/FIX/RECURSE cycle. Enforces fix-all-per-cycle rule — every confirmed finding must be fixed before moving to the next cycle. Use for quality audit, code audit, codebase review, technical debt audit, refactoring opportunities, module quality check, or architecture review.

Encodes a continuous improvement loop for goal-seeking agents: EVAL, ANALYZE, RESEARCH (hypothesis + evidence + counter-arguments), IMPROVE, RE-EVAL, DECIDE. Auto-commits improvements (+2% net, no regression >5%) and reverts failures. Works with all 4 SDK implementations. Auto-activates on "improve agent", "self-improving loop", "agent eval loop", "benchmark agents", "run improvement cycle".

Intelligent test selection based on code changes. Maps source files to tests via import analysis, implements tiered testing (fast < 1 min, impacted < 5 min, full suite), and tracks test reliability. Use when running tests after code changes to optimize feedback loops and CI time.

TLA+ formal specification expert for writing specs, model checking, and applying formal methods to amplihack workflows

Convert and browse session transcripts as HTML or Markdown. Supports Claude Code JSONL logs (auto-saved to ~/.claude/projects/) and GitHub Copilot CLI JSONL logs (auto-saved to ~/.copilot/session-state/*/events.jsonl). Auto-detects log source based on available directories and file format. Supports viewing the current session, a specific session by ID, agent background task output files, or all project sessions with optional date-range filtering.

Development workflow for features, bugs, refactoring. Normally executed as a sub-recipe by dev-orchestrator/smart-orchestrator. Supports direct invocation via recipe runner for standalone use.

6-phase investigation workflow for understanding existing systems. Normally executed as a sub-recipe by dev-orchestrator/smart-orchestrator. Supports direct invocation via recipe runner for standalone use.

Create, edit, and analyze Word documents with tracked changes support

Generates comprehensive GitHub activity reports across all authenticated accounts. Gathers repos, PRs, features, and themes for configurable time periods (1/5/7/30/90 days). Produces shareable markdown with tables, mermaid charts, and executive summaries. Can create a private repo with GitHub Actions automation and GitHub Pages aggregation site. Use when: "github report", "work report", "activity summary", "what did I work on", "gh-work-report", "show my github activity".

Expert project manager orchestrating backlog-curator, work-delegator, workstream-coordinator, and roadmap-strategist sub-skills. Coordinates complex software projects through delegation and strategic oversight. Activates when managing projects, coordinating work, or tracking overall progress.

Presentation creation, editing, and analysis. When Claude needs to work with presentations (.pptx files) for: (1) Creating new presentations, (2) Modifying or editing content, (3) Working with layouts, (4) Adding comments or speaker notes, or any other presentation tasks

Creates, refines, and validates Agent Skills following the open standard at agentskills.io, Claude Code extensions, and Anthropic best practices. Use when building, creating, generating, or designing new skills.

Convert and browse session transcripts as HTML or Markdown. Supports Claude Code JSONL logs (auto-saved to ~/.claude/projects/) and GitHub Copilot CLI JSONL logs (auto-saved to ~/.copilot/session-state/*/events.jsonl). Auto-detects log source based on available directories and file format. Supports viewing the current session, a specific session by ID, agent background task output files, or all project sessions with optional date-range filtering.

Comprehensive knowledge of amplihack framework architecture, patterns, and usage

Builds comprehensive, living code-atlases as multi-layer architecture documents derived from code-first truth. Defaults to both Graphviz DOT and Mermaid. User can override to single format. Language-agnostic (Go, TypeScript, Python, .NET, Rust, Java). Files issues with 'code-atlas-bughunt' label. Treats atlas-building as a multi-agent bug-hunting journey: graph-form reasoning exposes structural bugs, route/DTO mismatches, orphaned env vars, dead code paths, and stale documentation that linear review misses. Three-pass bug hunt with per-journey PASS/FAIL/NEEDS_ATTENTION verdicts. Use when: creating architecture documentation, investigating unfamiliar codebases, hunting structural bugs, setting up CI/CD diagram refresh, or publishing to GitHub Pages/mkdocs.

Autonomous co-pilot — agent formulates goal from natural language, enables lock mode with SessionCopilot reasoning, works until goal is achieved.

Fleet orchestration for distributed coding agents across Azure VMs. Invoked as `/fleet <command>`. Covers all fleet operations: status, scout, advance, adopt, watch, snapshot, dry-run, start, add-task, queue, auth, dashboard, tui, and more. Use when: user mentions fleet, agents, VMs, sessions, or asks "what are my agents doing".

Adopt GitHub Agentic Workflows (gh-aw) in any repository. Investigates existing workflows, identifies gaps, creates agentic workflows, and handles CI/merge issues. Auto-activates for gh-aw adoption, agentic workflow setup, or workflow automation requests.

Checks whether a pull request satisfies the project's merge criteria and records the required evidence in the PR description. Use with `/merge-ready` before review or merge when QA-team scenarios, docs links, quality-audit convergence, CI status, and diff scope must be verified.

Access Microsoft 365 data (emails, calendar, files) from WSL via the Windows Copilot CLI bridge. Routes queries through powershell.exe to the Windows-native Copilot CLI with WorkIQ plugin for M365 Graph API access.

Create, edit, and analyze Word documents with tracked changes support

Generates comprehensive GitHub activity reports across all authenticated accounts. Gathers repos, PRs, features, and themes for configurable time periods (1/5/7/30/90 days). Produces shareable markdown with tables, mermaid charts, and executive summaries. Can create a private repo with GitHub Actions automation and GitHub Pages aggregation site. Use when: "github report", "work report", "activity summary", "what did I work on", "gh-work-report", "show my github activity".

Presentation creation, editing, and analysis. When Claude needs to work with presentations (.pptx files) for: (1) Creating new presentations, (2) Modifying or editing content, (3) Working with layouts, (4) Adding comments or speaker notes, or any other presentation tasks

Convert and browse session transcripts as HTML or Markdown. Supports Claude Code JSONL logs (auto-saved to ~/.claude/projects/) and GitHub Copilot CLI JSONL logs (auto-saved to ~/.copilot/session-state/*/events.jsonl). Auto-detects log source based on available directories and file format. Supports viewing the current session, a specific session by ID, agent background task output files, or all project sessions with optional date-range filtering.

TLA+ formal specification expert for writing specs, model checking, and applying formal methods to amplihack workflows

Gherkin/BDD specification expert for writing behavioral scenarios, acceptance criteria, and applying structured specifications to improve code generation quality

Iterative codebase quality audit with multi-agent validation and escalating-depth SEEK/VALIDATE/FIX/RECURSE cycle. Enforces fix-all-per-cycle rule — every confirmed finding must be fixed before moving to the next cycle. Use for quality audit, code audit, codebase review, technical debt audit, refactoring opportunities, module quality check, or architecture review.

QA team for outside-in validation, side-by-side parity loops, and A/B behavioral comparison. Use when you need behavior-driven tests, legacy-vs-new comparison, or rollout shadow validation. Creates executable scenarios and parity workflows that agents can observe, compare, and iterate on. Supports local, observable tmux, remote SSH, and shadow-mode divergence logging patterns.