Workflow plugins

Teach your coding agent to use the Open Knowledge Format, portable structured knowledge for AI agents

1 skill

Start issue-to-pr runs from GitHub labels on your own machine: a scheduled tick picks the issue you labelled agent, or the one you replied on, and runs it headless under your own Claude Code or Codex login, on Windows, macOS or Linux.

1 skill

Validation hooks - protected branch guards, completion validation, anti-pattern detection

hooks

Render an implementation plan, tech spec, or codebase as an interactive architecture graph you edit directly; the structured edits flow back to the agent. Skill: /archeyes:review (needs `npx archeyes` — install `archeyes` from npm).

1 skill

Run owner-approved GitHub issues locally through Claude Code or Codex, with optional verified SEO-incident intake.

1 skill

3 lean agents across 6 phases with compound discovery loop, proactive research, and strict evaluation — RESEARCH, DISCOVER, BUILD, AUDIT, SHIP, LEARN

33 skills · 34 commands · 113 agents

Skill-driven workflow plugin for GitHub development with excellence-by-default quality gates. Encodes team knowledge as composable skills, enforces safety through hooks, compounds learning across sessions, and provides durable goals, workflows, triggers, and runs at .flow/. Requires bash, git, gh, jq, and python3 with PyYAML; on Windows, Git Bash (see docs/windows-support.md).

33 skills · 23 commands · 10 agents · hooks

Multi-phase research workflow for building cited, audited knowledge bases.

21 skills · 10 agents

An agent-engineering plugin for Claude Code: twenty-five fg-* skills. fg-security audits a codebase for exploitable vulnerabilities (methodology vendored from cloudflare/security-audit-skill, MIT) with artefacts kept outside the repo, and routes severity-gated findings into fix-forward plans. fg-debug diagnoses hard bugs conversationally (vendored from mattpocock/skills' diagnosing-bugs, MIT), then stops at diagnosis. It routes by state before size: an active failed task returns to fg-run even when trivial; only independent diagnoses branch to fg-quick or an approved fix-forward plan. A red command counts as a persistent eval only when durable; raw captures stay outside the repo and every exit cleans temporary artefacts. Unattended drives always skip it (ADR 260907-140655). Three form a loop that takes one task through a single cycle of ask·plan → execute → done: ask·plan (fg-ask) is grill-with-docs-style conversational grilling, execute (fg-run) picks a task from the backlog (Run all supported) and runs it as a Dynamic Workflow, and seal (fg-done) decides the retro — writing a retro log when there is something to learn and asking once only when a promotion to docs or evals clears the bar — then seals the task and guards against re-runs, via a deterministic seal script (forge-done.sh/.js) shared by every seal path (ADR-0030). Twenty-two more sit outside the loop: fg-learn batch-promotes the learnings unattended seals deferred; fg-map maps the codebase into .forge/codebase/ with parallel subagents so grilling reads a map instead of re-exploring the code (cuts context rot); fg-quick is a lightweight lane for trivial tasks that grills lightly, logs one line to .forge/quick/LOG.md, runs directly without formal artifacts (no ADR/plan/retro), and bails to fg-ask if the task turns out non-trivial; fg-status is a read-only reporter that surveys .forge/ and prints where every task stands plus the single next step you need (writes nothing, never auto-runs); fg-next derives th

26 skills · hooks

Generic GitHub workflow commands for issue management, PR creation, code review, and releases. Works with any repository by auto-detecting settings.

7 skills · 14 commands · 4 agents

Domain-driven build workflow: vision grounding, EventStorming, bounded-context mapping, and a dependency-ordered task backlog driven to done by TDD workers.

12 skills · 9 agents

Design-system authoring workflow: reference-grounded Socratic foundation, themable token sets (light+dark), and pure HTML/CSS Tailwind component kitchen-sinks with accessibility and i18n baked in.

12 skills · 6 agents

AI delivery pipeline for Claude Code — /evyasys:Docs, /evyasys:Plan, /evyasys:Build, /evyasys:Qa (plus Setup, Update, Repair, Diagnose). Plans work, builds with verified checks, tests without touching production, documents your code.

8 skills · 8 commands

Autonomous agentic loop that decomposes user queries into tasks, writes tests first, implements with fresh sub-agents, self-debugs on failure, and learns over time. Ships the super-ralph skill, /ralph and /super-ralph commands, and all ralph-* agents.

1 skill · 2 commands · 10 agents

Gated, evidence-first engineering process: classify a task, route it through spec, plan, implement, and cross-provider review with artifact gates.

3 skills · 4 commands · 3 agents · hooks

ocul-pm 훅 브리지(세션 신호) + oculpm-mcp(journal/plan 구조화 도구) + 스킬 5종(기록 규격·인셉션·self-audit·run-evals·tdd). .oculpm 추적 프로젝트에서만 동작합니다. macOS · Linux · Windows(베타 — Windows 훅은 Git Bash 필요, Windows 의 MCP 는 앱의 MCP 등록으로).

5 skills · 1 MCP server · hooks

Takes one feature from idea to reviewed pull request — spec, plan, build, review, ship — stopping to ask you at every step that leaves your machine. Requires spec-kit in the target repository.

4 skills · 1 command

Verdict first, receipts attached. No sycophancy, no caving, no unverified "done".

5 skills

Find, install (with consent) and enable MCP servers in the middle of a session. Tool pinning and quarantine, keychain secrets, audit log, native Claude Code /discover panel. Dashboard on :3424.

2 skills · 1 MCP server

Warns before a session runs out of context, then writes a handoff document the next session resumes from in a single file read.

2 skills · hooks

Runs a team's plan through the pipeline: set up the team once, then each member starts their next task with one command, and progress is kept for the team leader.

3 skills

Grade-gated /workflows: research -> design -> impl -> ci -> review -> testing. Every phase is scored against a rubric before it advances.

1 MCP server · hooks

Claude Code skills for cloud-native remote sensing and geospatial analysis

2 skills

A universal engineering judgment skill that helps Claude think like an experienced software engineer by understanding before editing, minimizing diffs, verifying changes, and following existing codebase conventions.

1 skill

Generates a scrum-tool-agnostic onboarding backlog (Jira, GitHub, or Azure DevOps) that takes a team from zero to publishing consumer + provider contract tests on PactFlow

1 skill

make it ask before it acts. — flip the helpful-assistant reflex: ask one sharp clarifying question instead of guessing.

1 skill

pull the whole routine. — chain tricks into one pipeline so the output of one flows into the next: `combo "frisk | launder | deadpan"`.

1 skill

Harness-native Cavaleiro Agent Harness operator layer - 71 agents, 311 skills, 97 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses

200 skills · 97 commands · 71 agents · 1 MCP server · hooks

Agents, skills, commands, hooks, and rules for PHP — PSR standards, PHPUnit, Pest, PHPStan, Psalm, Composer, PHP-CS-Fixer

24 skills · 18 commands · 10 agents · hooks

Plan → Confirm → Implement → Verify discipline for Claude Code. Tiered session boot, deterministic hooks (quality gate, protect-changes, branch-protect, secret-scan, conventional-commit, loop-detect, and more), 37 user-invocable skills across audit / workflow / meta categories, 6 review agents (code-reviewer, security-reviewer, qa-reviewer, planner, dead-code-remover, devils-advocate), a CLAUDE.md ruleset, lessons-as-memory, ADR decisions, plus the optional wiki and HTML-artifacts modules of the `cck` CLI.

40 skills · 6 agents · hooks