feat(ai): make adaptive orchestration the planning default

CHANGELOG.md

@@ -53,6 +53,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Changed
 
+- Production planning-analysis requests now explicitly select contextual-orchestrator
+  `auto` mode, delegating route/conduct execution policy to the shared
+  orchestration service without weakening ScopeWeave's authenticated,
+  fail-closed transport and response-boundary controls.
 - Accepted XML whitespace before exact Microsoft Project element delimiters
   while preserving the linear, regex-free import scanner and rejecting
   attributes, longer names, non-XML whitespace, nested unmatched blocks, and

docs/doctoring/contextual-orchestrator-auto-default.md

@@ -0,0 +1,37 @@
+# Contextual-orchestrator adaptive planning default
+
+## Status
+
+Active pull-request evidence. This record does not describe protected `develop` until the pull request is integrated.
+
+## Decision boundary
+
+ScopeWeave owns the meaning, authorization, and presentation of a planning-analysis request. The shared `contextual-orchestrator` service owns provider/model selection and the depth/topology of execution. Production ScopeWeave requests therefore send `orchestration_mode: "auto"` explicitly instead of relying on an implicit gateway default or selecting `route`/`conduct` locally.
+
+The binding dependency evidence verified for this slice is protected `ContextualWisdomLab/contextual-orchestrator` `main` commit `6841b71935e0b7cb98fb52bcb4709cc5100c8d87`. At that revision, the `/v1/chat/completions` contract accepts `orchestration_mode` and permits `auto`, `route`, and `conduct`.
+
+This decision does **not** promise a specific provider, model, worker count, topology, verifier strategy, or cost heuristic. Those remain shared-service policy and may evolve behind its versioned contract.
+
+## Security and standalone behavior
+
+The change preserves the protected ScopeWeave orchestrator boundary: authenticated canonical provider origin, HTTPS outside explicit loopback development, bounded messages, 120-second request timeout, bounded streamed provider responses, sanitized failures, and deterministic text only under explicit `SCOPEWEAVE_DEV=1` development mode. No provider credential or caller-controlled execution policy is added.
+
+## TDD and branch-safety evidence
+
+The feature branch originally contained a self-modifying write-capable workflow on a tree that had fallen behind protected `develop`. Continuing from it would have replaced later protected orchestrator hardening. That control path was removed and the branch was reconciled non-destructively with protected `develop@28420da358f57be5e85be3660251e39b85e1cc94` before product work resumed.
+
+On the reconciled tree, test commit `177e2a12e1f6be32c5a9b91379f92cf1170a71c8` first required `orchestration_mode: "auto"` in the existing production transport regression. Source commit `5a6afdd9f26994987e3d9ba3b85a8dd95f06a7e7` then added the field while preserving the existing fail-closed request boundary. Exact-current-head hosted evidence remains authoritative for merge readiness.
+
+## Rollback
+
+Rollback removes the explicit `orchestration_mode` request field, its regression expectation, this evidence record, the matching production-contract text, and the changelog entry together. Rollback must not restore the discarded self-modifying workflow or stale pre-hardening orchestrator source.
+
+## APA 7th references
+
+Contextual Wisdom Lab. (2026). *contextual-orchestrator* (Commit 6841b71935e0b7cb98fb52bcb4709cc5100c8d87) [Computer software]. GitHub.
+
+Nielsen, S., Cetin, E., Schwendeman, P., Sun, Q., Xu, J., & Tang, Y. (2025). *Learning to orchestrate agents in natural language with the Conductor*. arXiv. https://doi.org/10.48550/arXiv.2512.04388
+
+Sakana AI. (2026). *Sakana Fugu: Multi-agent system as a model*. https://sakana.ai/fugu/
+
+Xu, J., Sun, Q., Schwendeman, P., Nielsen, S., Cetin, E., & Tang, Y. (2025). *TRINITY: An evolved LLM coordinator*. arXiv. https://doi.org/10.48550/arXiv.2512.04695

docs/orchestrator-production.md

@@ -38,6 +38,17 @@ endpoint is absent. That variable must never be set in staging or production.
 
 ## Orchestration responsibility
 
+ScopeWeave explicitly sends `orchestration_mode: "auto"` together with the
+configured model and validated messages on production briefing requests. The
+current protected `ContextualWisdomLab/contextual-orchestrator` `main` contract
+verified for this change, commit
+`6841b71935e0b7cb98fb52bcb4709cc5100c8d87`, accepts `auto`, `route`, and
+`conduct` as orchestration modes. ScopeWeave chooses `auto` as its default so
+execution policy can be optimized centrally without coupling this product to a
+specific provider, worker count, topology, verifier pattern, or cost heuristic.
+Those internal choices remain `contextual-orchestrator` authority and are not a
+ScopeWeave compatibility promise.
+
 ScopeWeave intentionally sends only a versioned OpenAI-compatible request to
 the orchestration service. Model selection, single-model versus multi-agent
 allocation, task decomposition, role-specific reasoning effort, recursion

server/orchestrator.mjs

@@ -303,7 +303,11 @@ export async function chat(messages) {
         'content-type': 'application/json',
         authorization: `Bearer ${configuration.token}`,
       },
-      body: JSON.stringify({ model: OC_MODEL, messages: safeMessages }),
+      body: JSON.stringify({
+        model: OC_MODEL,
+        orchestration_mode: 'auto',
+        messages: safeMessages,
+      }),
       signal: AbortSignal.timeout(ORCHESTRATOR_TIMEOUT_MS),
     });
   } catch {

tests/unit/orchestrator.test.mjs

@@ -98,6 +98,7 @@ try {
   assert.ok(calls[0].init.signal instanceof AbortSignal);
   assert.deepEqual(JSON.parse(calls[0].init.body), {
     model: 'nvidia/nemotron-3-super-120b-a12b',
+    orchestration_mode: 'auto',
     messages: [{ role: 'user', content: 'status' }],
   });