dm-work

Generate a second-opinion brief for an external agent with no prior context

Review and align steering files (CLAUDE.md, AGENTS.md) with Claude Opus 4.6 prompting best practices

Decompose an epic or external specification into implementable task beads with complexity-based labeling.

Spawn N agents on the same spec in isolated worktrees, compare by objective metrics, pick winner

Compress a document or artifact for agent-optimized token density while preserving effectiveness

Pre-merge checklist for worktree branches. Ensures quality gates, review, and beads are complete before merging.

Run autonomous review and evaluation after a merge. Surfaces findings as beads for triage in the next session.

Transform an ambiguous specification into an implementable one through 4 adversarial passes.

Run parallel architecture, code, security, and design review with local beads or GitHub PR comments

Rotate the session — generate snapshot, save to disk, and prepare for /clear

Delegate implementation work to a subagent with appropriate profile, skills, and boundaries

Orchestrate multiple subagents in parallel or serial based on task dependencies and file ownership

Triage PR review comments - accept to create beads, reject with optional reply

Use when the user wants to explore approaches or discuss design before implementation. Explores intent, requirements, and design through collaborative dialogue.

QA web applications using Chrome DevTools MCP. Use when testing running apps, verifying acceptance criteria against a live UI, checking for console errors, evaluating UI behavior, or running regression checks. Requires the chrome-devtools MCP server to be connected. Complements dm-work:review (code-level) with runtime verification.

Power CLI tools (fd, rg, jq, yq, sd, xargs, bat, delta) for when built-in tools are insufficient. Use for complex file ops, data manipulation, or parallel execution.

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes. Requires root cause investigation before any fix attempts. Random fixes waste time and create new bugs.

Transform ambiguous specs into implementable work items through adversarial refinement. Use when receiving specs, claiming epics, or starting complex (l/xl) tasks. Use /breakdown for spec→tasks decomposition, /refine for sharpening individual items.

Grade implementation work against bead acceptance criteria using a separate judge agent. Use after subagent work passes mechanical gates, as a pre-merge check, or on-demand to evaluate existing features. The evaluator is NOT the orchestrator and NOT the implementer — it only judges. Integrates with browser-qa for runtime verification when CDT MCP is available.

Modern dev tool version management with mise. Use when setting up repos, managing tool versions, or when users ask about nvm/pyenv/goenv alternatives.

Activate at session start when you are the primary Claude instance. Establishes orchestrator role with delegation protocols, subagent launch templates, token efficiency rules, and parallel safety constraints. You orchestrate; subagents implement.

Reduce CLI output noise before it reaches LLM context. Use when setting up repos, troubleshooting context overflow from build/test/git output, or when Bash tool output exceeds limits. Covers RTK (zero-config baseline) and tokf (per-project customization). Either or both may be installed — the skill adapts gracefully when one is absent.

Analyze a repository's configuration against CC and workflow best practices. Use when auditing a repo, after repo-init, or when troubleshooting token usage or workflow gaps. Reports issues by severity with auto-fix offers.

Initialize a new repository with standard scaffolding - git, gitignore, AGENTS.md, justfile, mise, beads, and timbers. Use when starting a new project or setting up an existing repo for Claude Code workflows.

Use at end of session before committing, when landing the plane, or when user says "retro", "what did we learn", "session review". Lightweight self-improvement pass that turns session friction into persistent rules and memories. Runs inline — no separate documents, no approval prompts.

Activate when you are a delegated subagent (not the orchestrator). Establishes subagent protocol with terse returns, details to history/, file ownership boundaries, and escalation rules. You implement; orchestrator reviews and commits.

Use when implementing features or fixing bugs — write the test first, watch it fail, write minimal code to pass. Ensures tests verify behavior by requiring failure first.

Use when creating a worktree, setting up a worktree, starting feature work that needs isolation, or before executing implementation plans. Covers git worktree creation under .worktrees/, gitignore setup, beads integration, and merge guardrails.

5 hooks across 4 events