skill-orchestratorlisted

Plan and execute multi-skill CometWeb workflows with CW-AIP handoffs when the user wants one entry point instead of tagging each specialist. Use when the user asks to orchestrate, sequence, or run end-to-end flows (e.g. evidence then Council, audit then release, research then weekly ops), run everything needed for a goal, or chain Evidence Researcher with ai-council or other skills in order. Supports execution_mode auto|single_thread|isolated_subagents. Loads each step's SKILL.md, runs it fully, and passes envelopes between steps. Do not use for a single-domain task when one specialist skill is enough, for routing-only questions without execution, or to bypass AI Council decision policy or Release Readiness gate rules.
MaciejZet/agent-skills · ★ 1 · AI & Automation · score 74

Install: claude install-skill MaciejZet/agent-skills

# Skill Orchestrator Single entry point for **multi-skill workflows**. You plan the sequence, choose an execution mode, run each specialist skill, and hand off via CW-AIP — without collapsing research into decisions or skipping skill boundaries. ## Execution mode | Mode | Behavior | | --- | --- | | `auto` | Prefer `isolated_subagents` when the host supports Task/subagents **and** the workflow benefits from isolation (Council + research, audit + release, ≥2 heavy specialists). Otherwise `single_thread`. | | `single_thread` | Parent loads each `SKILL.md` and executes steps sequentially in this thread. | | `isolated_subagents` | One subagent per step; parent only plans, launches, validates envelopes, and merges. | `skill-orchestrator-multiagent` is a **thin alias** for `execution_mode=isolated_subagents`. Do not maintain a divergent planning model there. Optional context preflight: when the goal depends on current multi-source CometWeb state, run `cometweb-context` first and pass the ContextEnvelope as a dependency. ## Boundaries **Owns:** - workflow archetype selection, - `execution_mode` selection, - ordered execution plan with envelope types, - loading each step's `SKILL.md` (or launching a subagent that does), - handoff summaries between steps, - final workflow result (artifacts produced, gaps, what remains). **Does not own:** - Evidence Pack construction (Evidence Researcher), - GO/NO-GO / Council deliberation (AI Council), - release gate verdicts (Release Readine