tangle-network · drewstone · Jul 29, 2026 · Jul 29, 2026 · Jul 29, 2026 · Jul 29, 2026
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,5 +1,13 @@
 # Changelog
 
+## 0.110.0
+
+- Run every supervisor, including the root, from one complete `AgentProfile`, preserve exact profile/task/candidate identity through recursive delegation, and reject execution paths that would silently drop profile fields.
+- Expose node-scoped product tools, product authorization for exact spawns and continuations, awaited replay-safe coordination observation, structured worker traces, trace-derived failure guidance, and caller cancellation across the complete recursive run.
+- Make durable run and assignment identity stable across restart while retaining exact materialization, accounting, delivery, and settlement evidence for each node.
+- Add live root-manager steering, trusted post-authorization manager/leaf classification, per-assignment completion checks, a cold recursive forest reader, and public exact-profile candidate conversion helpers.
+- Align Runtime with Eval 0.135.4, Interface 0.37.0, Knowledge 7.0.0, and Materialize 0.9.3 so public-source provenance and exact knowledge claims use one package cohort.
+
 ## 0.109.2
 
 - Align Runtime with Eval 0.135.2 and Knowledge 6.1.11 so every improvement path uses the corrected paired promotion decisions.

diff --git a/README.md b/README.md
@@ -116,7 +116,13 @@ One supervisor spawns and steers workers toward a goal. Where the workers run (a
 import { supervise } from '@tangle-network/agent-runtime/kernel'
 
 const result = await supervise(
-  { name: 'supervisor', harness: null, systemPrompt: 'Delegate to workers; do not solve the task yourself.' },
+  {
+    name: 'supervisor',
+    harness: 'cli-base',
+    prompt: {
+      systemPrompt: 'Delegate to workers; do not solve the task yourself.',
+    },
+  },
   'Implement the feature and make the tests pass.',
   { budget, router, backend }, // backend = where workers run: router-tools | sandbox+harness | bridge
 )

diff --git a/docs/api/agent.md b/docs/api/agent.md
@@ -1106,16 +1106,73 @@ the loop produces 20 minutes later).
 
 AgentProfile axis name, with `custom:<name>` reserved for caller-owned extensions.
 
+***
+
+### CanonicalAgentProfileMaterializationAxis
+
+> **CanonicalAgentProfileMaterializationAxis** = [`KnownAgentProfileMaterializationAxis`](#knownagentprofilematerializationaxis)
+
+Canonical AgentProfile axes used when checking one complete profile.
+
 ## Variables
 
 ### AGENT\_PROFILE\_MATERIALIZATION\_AXES
 
-> `const` **AGENT\_PROFILE\_MATERIALIZATION\_AXES**: readonly \[`"identity"`, `"name"`, `"model"`, `"prompt"`, `"systemPrompt"`, `"instructions"`, `"resources"`, `"files"`, `"resourceInstructions"`, `"skills"`, `"resourceTools"`, `"resourceAgents"`, `"commands"`, `"tools"`, `"permissions"`, `"mcp"`, `"mcpConnections"`, `"connections"`, `"subagents"`, `"hooks"`, `"modes"`, `"confidential"`, `"metadata"`, `"extensions"`\]
+> `const` **AGENT\_PROFILE\_MATERIALIZATION\_AXES**: readonly \[`"identity"`, `"name"`, `"description"`, `"version"`, `"tags"`, `"model"`, `"modelDefault"`, `"modelSmall"`, `"modelProvider"`, `"modelReasoningEffort"`, `"modelMetadata"`, `"harness"`, `"prompt"`, `"systemPrompt"`, `"instructions"`, `"resources"`, `"files"`, `"resourceInstructions"`, `"skills"`, `"resourceTools"`, `"resourceAgents"`, `"commands"`, `"resourceFailOnError"`, `"tools"`, `"permissions"`, `"mcp"`, `"mcpConnections"`, `"connections"`, `"subagents"`, `"hooks"`, `"modes"`, `"confidential"`, `"metadata"`, `"extensions"`\]
 
 Known AgentProfile axes a run path may or may not carry into execution.
 
 ***
 
+### fullProfileMaterialization
+
+> `const` **fullProfileMaterialization**: [`ProfileMaterializationContract`](#profilematerializationcontract)
+
+Materialization contract for a run path that executes every canonical AgentProfile axis.
+
+***
+
+### promptModelProfileMaterialization
+
+> `const` **promptModelProfileMaterialization**: [`ProfileMaterializationContract`](#profilematerializationcontract)
+
+Materialization contract for an intentionally limited prompt-and-model execution path.
+Identity, harness, and metadata are control fields consumed for naming, placement,
+authorization, and durable attribution; they are carried without adding worker behavior.
+Every behavioral axis other than prompt and model remains unsupported.
+
+***
+
+### worktreeCliProfileMaterialization
+
+> `const` **worktreeCliProfileMaterialization**: [`ProfileMaterializationContract`](#profilematerializationcontract)
+
+Materialization contract for a local coding CLI in an isolated git worktree.
+The shared workspace materializer carries native tools, permissions, MCP, hooks, subagents,
+modes, and file-backed resources when the selected CLI supports their exact values. Runtime
+placement concerns (hub connections and confidential execution), provider-native extensions,
+unused model hints, and `resources.failOnError` are deliberately absent so they fail before a
+worktree or executor is created rather than being mistaken for an effective candidate change.
+
+***
+
+### controlProfileMaterialization
+
+> `const` **controlProfileMaterialization**: [`ProfileMaterializationContract`](#profilematerializationcontract)
+
+Materialization contract for a raw process path that carries only control/identity fields.
+
+***
+
+### promptControlProfileMaterialization
+
+> `const` **promptControlProfileMaterialization**: [`ProfileMaterializationContract`](#profilematerializationcontract)
+
+Materialization contract for an injected inference function whose surrounding driver still
+applies the profile prompt, name, placement, and metadata, but not model selection.
+
+***
+
 ### sandboxActProfileMaterialization
 
 > `const` **sandboxActProfileMaterialization**: [`ProfileMaterializationContract`](#profilematerializationcontract)
@@ -1275,6 +1332,28 @@ Define the profile axes a concrete run path actually carries into execution.
 
 ***
 
+### profileMaterializationAxes()
+
+> **profileMaterializationAxes**(`profile`): readonly (`"metadata"` \| `"resources"` \| `"name"` \| `"tools"` \| `"model"` \| `"mcp"` \| `"connections"` \| `"subagents"` \| `"hooks"` \| `"modes"` \| `"extensions"` \| `"description"` \| `"version"` \| `"tags"` \| `"prompt"` \| `"harness"` \| `"permissions"` \| `"confidential"` \| `"systemPrompt"` \| `"instructions"` \| `"skills"` \| `"identity"` \| `"modelDefault"` \| `"modelSmall"` \| `"modelProvider"` \| `"modelReasoningEffort"` \| `"modelMetadata"` \| `"files"` \| `"resourceInstructions"` \| `"resourceTools"` \| `"resourceAgents"` \| `"commands"` \| `"resourceFailOnError"` \| `"mcpConnections"`)[]
+
+Return the exact canonical axes a complete profile actually requests. Compound prompt, model,
+identity, and resource objects are split so a path cannot claim an entire object while silently
+dropping one of its fields.
+Empty strings, arrays, and nested records do not claim support; explicit
+scalar values such as `false` and `0` remain meaningful requests.
+
+#### Parameters
+
+##### profile
+
+`AgentProfile`
+
+#### Returns
+
+readonly (`"metadata"` \| `"resources"` \| `"name"` \| `"tools"` \| `"model"` \| `"mcp"` \| `"connections"` \| `"subagents"` \| `"hooks"` \| `"modes"` \| `"extensions"` \| `"description"` \| `"version"` \| `"tags"` \| `"prompt"` \| `"harness"` \| `"permissions"` \| `"confidential"` \| `"systemPrompt"` \| `"instructions"` \| `"skills"` \| `"identity"` \| `"modelDefault"` \| `"modelSmall"` \| `"modelProvider"` \| `"modelReasoningEffort"` \| `"modelMetadata"` \| `"files"` \| `"resourceInstructions"` \| `"resourceTools"` \| `"resourceAgents"` \| `"commands"` \| `"resourceFailOnError"` \| `"mcpConnections"`)[]
+
+***
+
 ### validateProfileMaterialization()
 
 > **validateProfileMaterialization**(`options`): readonly [`ProfileMaterializationIssue`](#profilematerializationissue)[]

diff --git a/docs/api/candidate-execution.md b/docs/api/candidate-execution.md
@@ -254,6 +254,12 @@ Re-exports [prepareAgentCandidateExecution](index.md#prepareagentcandidateexecut
 
 ***
 
+### agentCandidateProfileAsAgentProfile
+
+Re-exports [agentCandidateProfileAsAgentProfile](index.md#agentcandidateprofileasagentprofile)
+
+***
+
 ### applyExactAgentProfileDiff
 
 Re-exports [applyExactAgentProfileDiff](index.md#applyexactagentprofilediff)
@@ -266,6 +272,18 @@ Re-exports [assertCandidateProfileBinding](index.md#assertcandidateprofilebindin
 
 ***
 
+### freezeGenericAgentCandidateProfile
+
+Re-exports [freezeGenericAgentCandidateProfile](index.md#freezegenericagentcandidateprofile)
+
+***
+
+### omitUndefinedObjectFields
+
+Re-exports [omitUndefinedObjectFields](index.md#omitundefinedobjectfields)
+
+***
+
 ### parseExactAgentProfile
 
 Re-exports [parseExactAgentProfile](index.md#parseexactagentprofile)
@@ -278,6 +296,12 @@ Re-exports [parseExactAgentProfileDiff](index.md#parseexactagentprofilediff)
 
 ***
 
+### parseExactCandidateProfile
+
+Re-exports [parseExactCandidateProfile](index.md#parseexactcandidateprofile)
+
+***
+
 ### AgentCandidateModelGrantActivateInput
 
 Re-exports [AgentCandidateModelGrantActivateInput](index.md#agentcandidatemodelgrantactivateinput)