Run a multi-agent GAN-style design sprint to produce a frontend design specification. Three designer agents (sonnet) propose independently across three rounds of increasing concreteness; a persistent evaluator (opus) scores each round and issues a shared brief; the team lead writes the final spec. Use when the user…
Show a tree of epics with open/closed ticket counts. Use when the user asks to see an epic tree, visualize epic hierarchy, or show ticket counts by epic.
Implement a set of tk tickets in parallel using a DAG-driven agent team with continuous dispatching. Spawns a pool of up to 4 implementers and up to 2 quality reviewers at startup (sized to the input ticket count, so a single-ticket run gets 1 + 1); dispatches tickets as they unblock without wave boundaries. Quality…
Implement a set of tk tickets in parallel using an agent team. Analyzes tickets for file-level conflicts, groups into minimum-wave batches, and runs each wave with parallel implementers validated by a quality reviewer. Designed for multi-review fix batches where tickets don't have formal acceptance criteria. Use when…
Wire a shared opt-in rule from /.claude/optionalrules/ into the current project's .claude/rules/ as a symlink. Use when the user says "load the shared rule", "set up shared rule X for this project", "opt in to the skeletons rule", or similar. Takes one argument — the rule name (e.g. skeletons).
Perform a comprehensive code review using multiple specialized sub-agents in parallel. Covers logical correctness, performance, readability, and security. Creates tk tickets if tk is available, otherwise writes a report to .code-review/final-report.md.
Spawn a team of Playwright-driven agents to explore a running web app as simulated users. Supports ad-hoc exploration (routes discovered from source code) and catalog mode (predefined scenarios loaded by name). Uses wave-based execution with agent recycling to prevent context exhaustion. Findings become tk tickets.…
Quick code review using an adversarial sub-agent. Surfaces Critical and High issues first. Creates tk tickets for Critical/High/Medium findings if tk is available.
Execute a tk epic using a DAG-driven agent team with continuous dispatching. Spawns a pool of up to 4 implementers, up to 2 quality reviewers, and 1 AC verifier at startup (sized to the open ticket count, so a single-ticket epic gets 1 + 1 + 1); dispatches tickets as they unblock without wave boundaries. Use when the…
Execute a tk epic using an agent team. Creates implementer, AC verifier, and quality reviewer teammates, dispatches unblocked tickets for parallel implementation, and manages the validation loop. Use when the user says "run epic", "execute epic", "start working on the epic", or similar.
Scaffold a new Go project with a Makefile, golangci-lint, GitHub Actions CI, and two-layer git hooks (pre-commit + pre-push). Use when the user says "set up a new Go project", "scaffold a Go project", "create a new Go CLI", or similar.
Scaffold a new SvelteKit project with Biome, Prettier, Vitest, and GitHub Actions CI. Use when the user says "set up a new JS project", "scaffold a SvelteKit project", "create a new frontend app", or similar.
Scaffold a new Python project with uv, ruff, pytest, and GitHub Actions CI. Use when the user says "set up a new Python project", "scaffold a project", "create a new Python app", or similar.
Report a consistent status snapshot for an in-flight /run-epic, /fix-tickets, or DAG variant — the epic/batch progress, every ticket and its state, every agent and what it's doing, and any interesting events from the run so far. Use when the user asks "what's the status", "how's the run going", "what are the agents…
Triage, filter, and sort tk tickets for review. Use when the user asks to review open tickets, sort by priority/type/confidence, triage a backlog, go through P2 bugs, or any similar ticket review workflow. Always use tk triage instead of inline Python/jq for filtering and sorting.
Ship a completed /run-tickets, /run-epic or /fix-tickets batch: merge the integration branch to main, prune worktrees, move leftover non-blocking findings to the repo's backlog epic, close the epic with a ship note, and report remaining work. Use only when the user types /wrap-epic.
Analyze archived /run-tickets runs to find what to tune: rework and block rates, token cost, and escaped defects, broken down by effort policy, implementation and review effort, diff size, and dependents. Read-only; recommends changes with the evidence and sample size behind each. Use when the user asks how…
Control Herdr, a terminal multiplexer for coding agents. Use only when the user explicitly mentions Herdr or asks to use Herdr to inspect or control panes, tabs, workspaces, commands, or another agent. Do not use merely because a task could benefit from a background terminal, delegation, or parallel work. Requires…
Ticket-by-ticket status of a /run-tickets workflow run, running or finished: what each ticket is doing or how it ended, how long each step took, rework and conflict counts, findings filed, merges, and the integration check. Read-only. Use when the user asks how a /run-tickets run is going or went, how long it took…
Implement a set of tk tickets (an epic's children or an explicit list) in parallel using the run-tickets workflow: decisions put to the owner before dispatch, pooled worktrees, one fresh implementer per ticket, AC verification when the ticket has acceptance criteria, adversarial quality review, and gated one-at-a-time…
★not rated 5
changed 5d agoA85 tokens
originalMIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: