Workflow plugins

Work framework: workspace init (scoping interview → CLAUDE.md + master plan + skill routing), wave-based parallel execution on agent teams or subagents, ≤4h task breakdown with exit checks, mandatory decision briefs on every real fork, coordinator status blurbs, fresh-docs discipline for third-party dashboards.

6 skills · hooks

Implementation workflow: scoped subagents with a reviewing main thread, design principles, testing, research, plain reporting, and self-documenting code.

7 skills

Autonomous dev loop: plan, implement, codex review, claude review, fix, repeat. Orchestrates Claude + Codex for self-improving code with a Stop hook to prevent early exit.

1 skill · 3 agents · 1 MCP server · hooks

Clone your team's best code reviewers from their PR history (across one or more repos) and run them as a review panel. Extracts a persona per reviewer — what they care about, what they catch, how they sound — not a rule list.

1 skill · 3 commands · 2 agents

Turn a Slack bug-reports channel into Jira tickets — ledger-backed, idempotent, capped, with a threaded reply on every message it touches.

1 skill · 5 agents

Turn a Slack bug-reports channel into Jira tickets — ledger-backed, idempotent, capped, with a threaded reply on every message it touches.

1 skill · 5 agents

Spec-driven development workshop with beads wisps and thin stories: park, debug, shape, and build

3 skills · 1 agent · hooks

Your team's coding standards, in context. Connects over MCP to a dev-playbook server (run it locally with Docker) for guardrails, definition of done, per-language rules and task workflows - and scaffolds them for a repo that has none. Run /dev-playbook-init to set it up.

2 skills · 1 command · 1 MCP server · hooks

PRD and TECH_REQ generating commands that can be used pre-EPCC workflow

2 commands · 2 agents

Git shipping workflow: conventional commits with gitmojis, PRs and MRs opened and merged with the matching forge CLI, curated releases.

4 skills

Capture project sessions to a structured journal using a 6-field daily template, then harvest ⭐ items into permanent-note drafts grouped by theme

2 commands · 2 agents

Continuous self-referential AI loops for iterative development with session isolation fix

3 commands · hooks

Run an asynchronous review or decision loop between two AI sessions through one append-only Markdown file, with polling that survives restarts.

3 skills

Reflect on the current session and route each learning to the store that will actually enforce it — a deterministic hook, CLAUDE.md, a new skill, or memory.

3 skills

Execution planning and running. Generates validated execution plans from tech specs, executes them phase-by-phase with parallel task groups, and tracks progress in execution logs.

2 commands · 1 agent

Capture pages, publish OAuth-authenticated snapshots with the Screenote CLI, and retrieve visual feedback from Claude Code or Codex.

3 skills

Portable framing discussions — size the decision, then list assumptions, rank hypotheses, or interview, converging on an explicit decision. No external CLI dependency.

1 skill · 1 command

Reusable desktop UI/UX patterns (framework-agnostic) — ribbon, command bar, tabs, panels, etc.

1 skill · 1 command

Cross-agent behavioral guidelines to reduce common LLM coding mistakes: Think Before Coding, Simplicity First, Surgical Changes, Goal-Driven Execution

1 skill

Safe merge-to-main workflow with pre-merge checks, conventional commits, doc updates, and post-deploy verification.

1 skill

Spec-Driven Development (SDD) — a structured, specification-first workflow for planning and executing software features with Claude. Covers five phases: initialize, specify, plan, task, and execute. Includes Graphify integration for knowledge-graph-aware codebase navigation.

1 skill

Guidelines comportamentais para reduzir erros comuns de LLM em código: pensar antes de codar, simplicidade primeiro, mudanças cirúrgicas, execução orientada a critério verificável. Upstream de terceiro (MIT, forrestchang) — listado aqui só para distribuição, sem fork; sha pinado.

1 skill

Behavioral guidelines to reduce common LLM coding mistakes: Think Before Coding, Simplicity First, Surgical Changes, Goal-Driven Execution

1 skill

Documentation authoring — structured markdown with YAML frontmatter, Mermaid diagrams, and auto-fixed lint-clean output.

1 skill

Developer-focused dynamic workflows for Claude Code: codebase-survey, bug-sweep, and compare-options. Each fans out into a multi-agent run.

1 command

Optional Bash-tool and workflow enforcement hooks for the agent plugin: no-bash, no-absolute-paths, force-plan-mode, auto-formatter, no-honest, no-idle-wait, no-tmp. Install to opt into enforcement. Requires agent@ronin-skills.

hooks

Lets an agent read the review queue zen-review keeps for a branch and answer it, without the loop being explained to it first.

1 skill

Safe workflows for disposable remote coding-agent sessions with bivouac

1 skill

Behavioral guidelines to reduce common LLM coding mistakes: Think Before Coding, Simplicity First, Surgical Changes, Goal-Driven Execution.

1 skill

Authorize an agent via device flow, obtain an API key, pick a model by price, and configure the client for SubRouter

1 skill