name: research-routes-run description: Execute an approved Research Routes context across software, research-exploration, or hybrid journeys with concurrent route owners, shared control-plane resources, CounterPilot read-only boundaries, explicit evidence and provenance, lifecycle recovery, and HumanDecision seams. Use when implementation or experiments must proceed without treating AI inference as fact or silently changing canonical state. metadata: version: "1.0.0"

Research Routes Run

Execute the approved route graph through the shared control plane and preserve route identity.

• Load the context and validate it before dispatch.
• Keep independent route owners concurrent where dependencies permit; share only the shared backbone and declared resources.
• Run CounterPilot as read-only challenge work. Record its boundary and findings; never let it mutate nodes or promote a route.
• Require provenance for evidence and preserve all polarity, including failure, conflict, boundary, and unknown.
• Treat implementation as distinct from canonical. A HumanDecision is required before canonical promotion or setting current continuation.
• Use lifecycle events to rewind or recover; never rewrite or delete prior route history.
• Reject missing, cross-context, or ambiguous IDs rather than defaulting to GraphPE, a repository, or a preferred route.
• Route authorization is a Pack record bound to a confirmed HumanDecision; route authorization alone does not authorize implementation or canonical promotion.
• A canonical downgrade is an explicit append-only transition that keeps the previous canonical reference and its failure evidence.

Use the shared control-plane, resource, recovery, and router contracts under shared/ together with the existing All in Luna execution machinery. This overlay changes the semantic vocabulary and entry points, not the shared orchestration behavior.