Skip to content

Skills

This page indexes all shared skills in this library, grouped by top-level skill type.

Skills by Category

Agentic

  • compose-agents-md — Create project-specific AGENTS.md instructions from the agentic library.
  • configure-mcp — Set up and configure MCP servers for agent tool access.
  • deploy-config — Deploy composed config, vendor files, and skills to a target project.
  • github-issue-planning — Issue-backed plan persistence on GitHub (prefer MCP, fallback to gh CLI).
  • memory-continuity — Maintain concise cross-session handoff memory with explicit MEMORY/index templates, snapshot indexing, and multi-agent worktree coordination.
  • plan-memory-index — Keep indexed, concise memory entries for applied plans under .agentic/memories/plan-memory/, while keeping .agentic/plans/ as plan source of truth.
  • persist-plan — Save the current plan as a structured file under .agentic/plans/.
  • write-plan — Auto-persist generated plans and keep a local plans index updated.

Backend

  • ddd-aggregate-modeling — Model aggregates, value objects, invariants, snapshots, and repository boundaries in DDD services.
  • golang-manual-di — Implement explicit dependency injection wiring patterns for Go services.
  • health-check-endpoints — Add liveness/readiness health endpoints aligned with probe conventions.
  • microservices-architecture — Design microservice boundaries, communication, and resilience strategy.
  • nosql-database-design — Model NoSQL schemas from access patterns and consistency needs.
  • serverless-architecture — Design and implement serverless workloads with operational guardrails.
  • sql-query-optimization — Diagnose and optimize SQL performance using plans, indexes, and rewrites.
  • technical-roadmap-planning — Build prioritized engineering roadmaps with dependencies and milestones.
  • terraform-infrastructure — Structure Terraform repo/stack layout, state topology, CI flow, and multi-region composition.
  • webhook-development — Build secure, reliable webhook receivers/senders with retries and idempotency.

Data

  • json-to-toon — Convert JSON into TOON format to reduce LLM token usage.
  • relational-database-design — Design or review relational schemas, constraints, and migrations.

Development

  • add-tests — Add focused tests for existing code based on behavior and edge cases.
  • code-review — Orchestrates code review by detecting project language and architecture, routing to specialized language skills (Go, TypeScript, Python, PHP) and architecture skills dynamically.
  • code-review-go — Deep Go-specific code review: goroutine lifecycle, data races, error wrapping, domain modeling, and interface design.
  • code-review-typescript — Deep TypeScript-specific code review: type safety, ESM/CJS, async correctness, money precision, and domain modeling.
  • code-review-python — Deep Python-specific code review: type annotations, async pitfalls, mutable defaults, threading safety, and domain modeling.
  • code-review-php — Deep PHP-specific code review: strict types, PHPStan compliance, PSR standards, domain modeling, and ORM boundary violations.
  • code-review-architecture — Architecture-focused code review: hexagonal boundary violations, DDD anti-patterns, CQRS misuse, and microservices coupling issues.
  • project-map — Generate a compact .agentic/project-map.md orientation document so any agent can understand codebase structure without filesystem exploration.
  • fix-bug — Investigate root cause and implement a verified bug fix with regression coverage.
  • git-flow-pr — Execute branch, commit, rebase, push, and pull request workflow end-to-end.
  • git-worktree-workspaces — Use Git worktrees to run parallel branch workspaces safely.
  • github-actions-workflow — Create or improve GitHub Actions CI/CD workflows.
  • monorepo-management — Set up and govern monorepo layout, tooling, and CI strategy.
  • new-gh-issue-orchestration — Run an issue-driven workflow from intake to PR with reviewer-first delegation.
  • pull-request-automation — Improve PR process with templates, labels, checks, and branch policies.
  • refactor — Refactor code for clarity and structure without changing behavior.
  • static-code-analysis — Configure and integrate static analysis tooling into local and CI workflows.

Devops

  • create-terraform-module — Create AWS-oriented Terraform module scaffolding with explicit interfaces, provider constraints, and reusable layout guardrails.
  • create-terraform-tests — Design and implement risk-focused Terraform module tests (terraform test) with CI-friendly validation flow.
  • design-aws-terraform-iac — Plan AWS Terraform architecture, service/module boundaries, state strategy, and acceptance criteria before implementation.
  • terraform-risk-playbook — Diagnose Terraform/OpenTofu risk first, enforce response contract, version guards, validation, and rollback expectations.
  • docker-compose-local-setup — Configure local multi-service docker compose stacks with readiness, envs, volumes, migrations, and verification.
  • use-aws-mini-stack-emulator — Use lightweight AWS emulation safely for local Terraform feedback loops, with explicit handoff to real AWS checks.
  • write-dockerfile — Generate production-ready Dockerfiles with secure, efficient build patterns.

Documentation

  • sync-confluence — Sync engineering docs from the repo into Confluence Cloud.
  • format-markdown — Format Markdown files with repo-native tooling first, then validate with markdown checks/lint.
  • write-adr — Create Architecture Decision Records with context, decision, and consequences.
  • write-changelog — Generate release changelog entries from commit history.
  • write-readme — Generate or update README content based on project source and structure.

UI

  • internationalization-i18n — Add or improve i18n/l10n architecture, messages, and workflows.
  • next-application-structure — Define scalable Next.js App Router project structure and conventions.
  • react-application-structure — Define scalable React application architecture and folder layout.
  • react-component-design — Design maintainable React component APIs and composition boundaries.
  • react-component-testing — Apply layered React testing strategy with RTL and Playwright.
  • vue-application-structure — Define scalable Vue application architecture and conventions.
  • vue-component-design — Design robust Vue component APIs with clear contracts.
  • vue-component-testing — Apply layered Vue testing strategy for logic, components, and flows.
  • wireframe-prototyping — Produce low-fidelity wireframes with interaction and acceptance notes.
  • diagram-design — Design implementation-grounded diagrams and flag ADR/RFC/code mismatches.

Using a skill outside a profile

Some profiles do not include every skill by default. There are two supported paths:

  1. Persistent/additive path: add the skill to .agentic/profile.yaml, run agentic sync to update AGENTS composition, then run agentic deploy <profile> [target] <vendor> to refresh .agentic/skills/.
  2. Ad hoc selective path: pass explicit skills during deploy, for example agentic deploy <profile> <target> <vendor> --skills <skill-name>.

When --skills is used, it acts as a selective override for that deployment rather than deploying the full library skill set.