Menú
Orchestrate autonomous workflow runs with queueing, concurrency, timeout, and retry controls.
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Basado en la clasificación ocupacional SOC
Transform product requirements into a focused UI/UX design brief with screens, flows, visual direction, and optional image-generation concepts. Bridge between product thinking and implementation.
Stage 3 of idea-to-delivery: gated engineering workflow that receives PRD, roadmap, and architecture inputs, decomposes into stories, and executes with TDD, traceability, and review gates.
Planning and decomposition review skill adapted from ECC planner agent. Covers story-set review (Gate 2) and planning consults for upstream skills. PRD-quality review is NOT in scope — that is owned by `woos-product-prd-review-gate`.
Produce a lean story plan (table) for the Gate 3 execution loop — execution order plus diff scope per story. PRD AC is the spec; tests are the verification. No per-story prose documents.
Resolve multi-agent review conflicts with explicit authority matrix, evidence weighting, and escalation criteria.
Architecture review and design skill adapted from ECC architect agent. Use for design validation, trade-off analysis, and system-level decisions.
| name | woos-run-orchestrator |
| description | Orchestrate autonomous workflow runs with queueing, concurrency, timeout, and retry controls. |
| version | 1.5.0 |
| author | Hermes Profile |
| license | MIT |
Provide execution control plane for near-unattended workflow runs.
queue_policy: FIFO | prioritymax_concurrencytimeout_seconds per stageretry_policy (attempts + backoff)cancellation_policyreview_loop_policy (review_round_max=2, reconciliation_attempt_max=1, escalation action)review_context_policy (load/update enforcement at review gates)review_runtime_policy (max_review_runtime_seconds per reviewer invocation)agent_memory_policy (persisted reviewer context per run)run_manifest_policy (run_manifest_required=true, verifier hard-fail rules)Runtime root resolution:
hep_root = <workspace_root>/hep (canonical default)hep_root for discoverability.runs/ is NOT required.Required runtime directories:
<hep_root>/runs/<run_id>/<hep_root>/review-context/Required file: <hep_root>/runs/<run_id>/run-manifest.yaml
Required sections:
phase_ordergate_results (must include gate enforcement payloads)baseline_decisions (deviation status + ADR/approval refs when needed)escalation_countersfinal_statusIf file or required sections are missing, status MUST be BLOCKED.
<workspace_root>/hep when missing.REQUEST_CHANGES loops and trigger escalation.BLOCKED and triggers handoff.reconciliation_attempt_max exceeded -> BLOCKED + handoff.review_round_max exceeded -> BLOCKED + handoff.max_review_runtime_seconds exceeded -> BLOCKED + handoff.BLOCKED + handoff.status: PASS | REQUEST_CHANGES | BLOCKEDrun_idstage_statusestimeoutsretriesreview_loops_detectedescalationshandoff_triggeredrun_manifest_verified