git clone --depth 1 https://github.com/jeremylongshore/tons-of-skills-marketplace /tmp/dispatch && cp -r /tmp/dispatch/plugins/ai-agency/hyperflow/skills/dispatch ~/.claude/skills/dispatchSKILL.md
# Dispatch Workhorse phase. Picks up a task file from `/hyperflow:plan` and runs it through the orchestrator pattern with parallel worker dispatch and multi-level reviews. This skill exercises **Layer 3 (Orchestrator)**, **Layer 5 (Quality Gates)**, **Layer 6 (Project Memory)**, **Layer 8 (Git Workflow)**, and **Layer 9 (Security)** from the doctrine. Multi-level review (L1–L5) is applied per the triage's flow profile. ## Per-Step Agent Map (DOCTRINE rule 12 — §12.1 inline-allowed for trivial steps · §12.2 sub-phase decomposition) Every substantive step dispatches at least one Agent. Trivial steps (≤ 2 tool calls, no content generation, no decision-making, mechanically verifiable) MAY be performed inline by the orchestrator per §12.1. Non-trivial steps decompose into ≥ 2 named sub-phases per §12.2. | Step | Sub-phase | Workers | Reviewer | Notes | |---|---|---|---|---| | 0 — Mode confirm | — (exempt) | — | — | `AskUserQuestion` only | | 0.5 — Operational choices | — (exempt) | — | — | `AskUserQuestion` only | | 1 — Load task | — (atomic · §12.2.8) | — | — | Read + schema check = single mechanical decision; no parallel angles | | 2a — Pre-dispatch | Composer × N parallel — one per sub-task; stitches persona + injects learnings | **Reviewer** — reviews prompt set for completeness | Parallel worker prompts built before any fan-out fires | | 2b — Worker fan-out | Implementer / Searcher / Writer × N parallel | **Domain specialist Reviewer** — the `Specialist:`-matched agent, batched over full batch (P2) or per-sub-task fallback | One Reviewer call per batch; security/correctness specialists run with `--thorough` | | 2c — Gate run | Worker — runs lint/typecheck/tests on affected files | **Reviewer** — judges gate output | Small focused diff | | 2d — Learnings + commit | Writer — synthesizes per-batch learnings | — (mechanical commit · §12.1) | Per-sub-task PASS commits land here; learnings appended to context | | 3 — Final integration review | — (atomic · §12.2.8) | **Reviewer** — broadest matching specialist(s), L1–L<n> over full diff | Single Reviewer dispatch; skipped under D7 incl. single-specialist coverage (rule 17) | | 4 — Wrap up | Writer — optional; only if memory prose is non-trivial | — | §12.1 trivial-inline; no Reviewer (D5) | | 5 — End of chain | — (exempt) | — | ONE `AskUserQuestion` with audit + deploy questions | Iron rule — `review agents ≥ batches + 1` (one batched Reviewer per batch + final integration when not skipped). The batched Reviewer counts as 1 per batch regardless of how many sub-tasks are in the batch. If less, a per-step reviewer was skipped. ## Review Levels (scale by flow profile) Every batch reviewer and the final integration reviewer uses the level set below. Profile comes from `/hyperflow:plan` triage and is propagated via the chain args (`triage=`). | Profile | Levels | Workers | Reviewers | |---|---|---|---| | `fast` | L1 | 1 | inline self-review only | | `standard` | **L1–L2 default** | 1–2 | 1 per-batch reviewer | | `deep` | L1–L5 | 3+ | per-batch + final integration | | `research` | L1–L2 + synthesis | 3+ searchers | inline synthesis | | `creative` | L1–L3 + UX | 1–2 | 1 reviewer | | `scientific` | L1–L5 + TDD | 2–3 | per-batch + final | L1 syntax/format · L2 spec/naming/edges · L3 integration/security · L4 perf/scale · L5 a11y/UX. See [review-levels.md](references/review-levels.md) for the full checklist. **Default cap is L1-L2.** Triage may flag `security: true` or `integration_risk: true` in its output; when either is set, the cap elevates to L1-L3 for both per-batch and final integration reviewers. Workers do NOT request elevation — only the upstream triage classification can elevate. See `reviewer-prompt-batched.md` — workers must honor the cap passed to them (cap enforcement lives on the reviewer-prompt side). ## Approval Gates | Gate | When | Format | |---|---|---| | Session context | Step 0, resolved (not asked) | inherited `session=` / handoff `HANDOFF.md` / default `one` | | Phase-dispatch scope | Step 1.5, feature mode with ≥ 2 incomplete phases | `AskUserQuestion` — all phases / phase by phase | | Inter-batch (manual mode only) | After each batch's gates pass | `AskUserQuestion` — continue / stop. **Auto mode fires NO inter-batch question** — see DOCTRINE rule 8 (invented gates banned). | | Hard halt | Any `SECURITY_VIOLATION` from a reviewer | Stop the chain, surface the finding | | **Audit prompt** | Step 5, after wrap-up | `AskUserQuestion` — run `/hyperflow:audit`? (yes/no, recommended toggles with flow profile) | | **Deploy prompt** | Step 5, after audit gate | `AskUserQuestion` — run `/hyperflow:deploy`? (yes/no, recommended toggles with gate state) | ## Inputs - **Task artefact** — positional arg (slug or path): either a flat `.hyperflow/tasks/<slug>.md` **or** a feature folder `.hyperflow/features/<slug>/` (see [`../hyperflow/feature-phases.md`](../hyperflow/feature-phases.md)). Default — the most-recently-modified of either. - **Handoff package** — a positional slug/path resolving to `.hyperflow-handoff/<slug>/` (see [`../hyperflow/session-handoff.md`](../hyperflow/session-handoff.md)). When present, dispatch is a **second-session build**: it rehydrates `artefact/` into `.hyperflow/` (Step 1.0) and reads `session`/`handoff`/chain args from `HANDOFF.md`. `on_complete` (review|deploy) governs Step 5. - **`session=<one|two>`** — passed in by `/hyperflow:plan` (or read from a handoff package's `HANDOFF.md`). If absent, assume `one`. In a two-session build, `handoff=<review|deploy>` governs the end-of-build behavior at Step 5. - **`--from-batch <n>`** — resume from a specific batch (skip prior batches). - **`--final-only`** — skip batch dispatch, run only the final integration review. - **`--thorough`** — disable P2 batched reviews; fall back to per-sub-task reviewers for every sub-task in every batch, and always run the final integration review (D7 skip is disabled). Use when belt-and-suspenders depth is required on a h
Guard the beads execution record: enforce the write-flush-verify discipline that defeats the bd rapid-write race, audit epic dependency graphs for cycles and orphans, catch closures whose title overstates what shipped, flag open beads carrying no disposition or a disproven premise, and reconcile bd against its GitHub and Plane projections. Owns RECORD INTEGRITY; delegates graph analysis to bead-dependency-mapper and epic-closure drift to bead-epic-auditor rather than duplicating them. Use before closing an epic, after any batch of bd writes, when a bead premise looks stale, or when auditing whether the record matches reality. Trigger with "audit beads", "check the bead DAG", "did that close actually land", "bead hygiene".
Verify every factual assertion in a diff, PR body, commit message, bead note, or governing doc against the actual repository, and fail anything that cannot be substantiated by a command. Use before merging any PR that makes claims about counts, coverage, consumers, enforcement, provenance, or certification, and when auditing standing docs for rot. Trigger with "verify claims", "check this PR body", "is this claim true", "claim audit".
Design and build Omarchy (Quickshell/QML) bar-widget, panel, and service plugins that actually work on a stock install. Knows the hard runtime constraint (no node on the graphical session PATH), the first-party contracts (BarWidget, Panel, KeyboardPanel, PanelKeyCatcher, Service), the curl-from-QML data pattern, FileView persistence, and the marketplace submission bar. Use when starting a new Omarchy plugin, porting a plugin off an external runtime, wiring a service to a bar widget, or deciding how a widget should fetch and persist. Trigger with "build an omarchy plugin", "omarchy widget", "quickshell plugin", "port this plugin to QML".
Audit an Omarchy plugin before it reaches the marketplace: prove it installs and runs on a stock box (no node/python on the session PATH), run the omarchy-submit gate lane, validate on the rig with omarchy-plugin-validate and qmllint, and check the QML security invariants and first-party idiom contracts. Read-only: it reports and blocks, it does not rewrite the plugin. Use before submitting an entry, after any data-layer change, or when a plugin works on the dev box and you need to know whether it works for a real user. Trigger with "audit this omarchy plugin", "is this plugin submission ready", "will this plugin work when installed".
Audit and fix Claude Code SKILL.md files against enterprise compliance standards: frontmatter completeness, required body sections, and style. Use when validating or repairing skills in a plugin directory. Trigger with "audit skill", "fix skill compliance".
Learn how SKILL.md files work in Claude Code plugins, then build a production-quality agent skill from scratch. Covers frontmatter schema, body structure, testing, and iteration.
Step-by-step guide to writing a SKILL.md file for Claude Code. Learn how to plan, structure, and test auto-activating skills with proper frontmatter, allowed-tools, dynamic context injection, and supporting files.
|