miadi-orchestration-kit

OpenClaw Runtime Patterns

RISE path: Reverse-engineer -> Intent-extract -> Specify -> Export

Control-plane scaffold for reconstructing OpenClaw runtime abstraction, durable workflow behavior, CLAW/gaia-endpoint topology, model-swap survival, session isolation, and runtime-inspector evidence.

This folder is created for jgwill/miadi-orchestration-kit#15 and should stay aligned with miadisabelle/workspace-openclaw#80.

Served tracks

Track	Search	Role
Academic	A1	Anchor runtime abstraction vocabulary and durable workflow claims.
Technical	T4	Map concrete runtime/session/endpoint patterns with path-level evidence.
Technical	T1/T3	Connect runtime choices to model routing and memory provenance without collapsing boundaries.

Inputs

Future waves should read these before adding findings:

1. OpenClaw model-routing study launch-manifest.md.
2. OpenClaw deep-research PROPOSAL.md, especially runtime abstraction and T4.
3. llms-rise-framework.txt.
4. OpenClaw/CLAW/gaia-endpoint repo path evidence before implementation claims.
5. Kherix or other external drafts only after ledger entry as provisional external research.

What is in this folder

File	Purpose
01-reverse-engineer.md	Reconstruct what source evidence actually shows before deriving intent.
02-intent.md	State desired outcome, current reality, structural tension, non-goals, and stop conditions.
03-specify.md	Define the reusable orchestration contract future waves must satisfy.
04-export.md	Provide launch, resume, audit, handoff, and promotion shapes for later work.
05-source-ledger.md	Preserve claim-level provenance, evidence type, status, contradictions, and reuse rules.

Acceptance criteria

• The exact RISE wording Reverse-engineer -> Intent-extract -> Specify -> Export appears in the README and is used as the folder’s stage language.
• A1 and T4 can share runtime vocabulary while preserving evidence type and confidence.
• Every runtime claim names whether it comes from transcript, study note, repo path, external draft, or direct inspection.
• Session isolation, endpoint maps, model-swap behavior, and workflow durability are tracked separately.
• Runtime-inspector needs are specified as read-only unless the manifest changes.
• No production runtime behavior is asserted without path-level or independently corroborated evidence.
• The scaffold does not run Academic, Technical, Narrative, or final research.

Source-ledger rules

• Evidence-backed claims need source surfaces and claim status.
• Repo path evidence is required before an implementation pattern is treated as observed behavior.
• Transcript-only claims are provisional until corroborated.
• Provisional claims must name the missing evidence needed for promotion.
• Contradictions stay visible until resolved; do not smooth them into consensus prose.
• Human-facing synthesis may only use claims cleared by 05-source-ledger.md or label them provisional.

Boundary

This rispec prepares future runtime research. It does not inspect live runtime state, run OpenClaw, operate gaia-endpoint, or implement the runtime inspector.