@poncho-ai/harness 0.51.1 → 0.52.1

package/.turbo/turbo-build.log

@@ -1,5 +1,5 @@
 
-> @poncho-ai/harness@0.51.1 build /home/runner/work/poncho-ai/poncho-ai/packages/harness
+> @poncho-ai/harness@0.52.1 build /home/runner/work/poncho-ai/poncho-ai/packages/harness
 > node scripts/embed-docs.js && tsup src/index.ts --format esm --dts
 
 [embed-docs] Generated poncho-docs.ts with 4 topics
@@ -8,9 +8,9 @@
 CLI tsup v8.5.1
 CLI Target: es2022
 ESM Build start
-ESM dist/index.js            535.75 KB
+ESM dist/index.js            536.24 KB
 ESM dist/isolate-F2PPSUL6.js 53.82 KB
-ESM ⚡️ Build success in 234ms
+ESM ⚡️ Build success in 235ms
 DTS Build start
-DTS ⚡️ Build success in 7701ms
-DTS dist/index.d.ts 91.75 KB
+DTS ⚡️ Build success in 8126ms
+DTS dist/index.d.ts 92.40 KB

package/CHANGELOG.md

@@ -1,5 +1,24 @@
 # @poncho-ai/harness
 
+## 0.52.1
+
+### Patch Changes
+
+- [`0e8fff1`](https://github.com/cesr/poncho-ai/commit/0e8fff12aed9d5efe1821ed3560ead48a16113c1) Thanks [@cesr](https://github.com/cesr)! - Only send `temperature` to the model when the agent explicitly sets one. The harness previously defaulted to `temperature: 0.2` and always passed it to `streamText`, which returns a 400 ("`temperature` is deprecated for this model") on models that removed sampling params (Fable 5, Opus 4.7+). `temperature` is now omitted from the request when undefined — the same treatment `maxTokens` already had — and `defaultAgentDefinition` no longer hard-codes a `temperature` line into the generated frontmatter (pass `temperature` explicitly to set one).
+
+## 0.52.0
+
+### Minor Changes
+
+- [`d8453b4`](https://github.com/cesr/poncho-ai/commit/d8453b4f2360a1734e448960fe52f6c450cdf842) Thanks [@cesr](https://github.com/cesr)! - harness: propagate `suppressTelemetry` to subagents.
+
+  A telemetry-off run (e.g. incognito) now suppresses telemetry for the subagents it spawns too, not just the parent turn. The parent run's `suppressTelemetry` is exposed on `ToolContext`, captured by `spawn_subagent` into the new `SubagentManager.spawn({ suppressTelemetry })` option, stored on the subagent conversation's `subagentMeta`, and read back by the orchestrator's `runSubagent` / continuation so the child run (and its re-runs) emit no `invoke_agent` / `execute_tool` / AI-SDK spans.
+
+### Patch Changes
+
+- Updated dependencies [[`d8453b4`](https://github.com/cesr/poncho-ai/commit/d8453b4f2360a1734e448960fe52f6c450cdf842)]:
+  - @poncho-ai/sdk@1.14.0
+
 ## 0.51.1
 
 ### Patch Changes

package/dist/index.d.ts

@@ -195,6 +195,10 @@ interface Conversation {
         status: "running" | "completed" | "error" | "stopped";
         result?: _poncho_ai_sdk.RunResult;
         error?: _poncho_ai_sdk.AgentFailure;
+        /** Inherited from the parent run at spawn time — when true, this
+         *  subagent's runs emit no telemetry (e.g. spawned from an incognito
+         *  turn). Read by the orchestrator's runSubagent / continuation. */
+        suppressTelemetry?: boolean;
     };
     channelMeta?: {
         platform: string;
@@ -733,7 +737,12 @@ interface DefaultAgentDefinitionOptions {
     modelProvider?: "anthropic" | "openai" | "openai-codex";
     /** Model name. Default: "claude-opus-4-5". */
     modelName?: string;
-    /** Sampling temperature. Default: 0.2. */
+    /**
+     * Sampling temperature. When unset, it is omitted from the generated
+     * frontmatter entirely and the harness sends no temperature (provider
+     * default). Newer models (Fable 5, Opus 4.7+) reject `temperature` — leave
+     * this unset for them.
+     */
     temperature?: number;
     /** Max tool-call steps per run. Default: 20. */
     maxSteps?: number;
@@ -1178,6 +1187,9 @@ interface SubagentManager {
         parentConversationId: string;
         ownerId: string;
         tenantId?: string | null;
+        /** Inherit the parent run's telemetry choice — when true, the subagent
+         *  run (and its re-runs) emit no telemetry. */
+        suppressTelemetry?: boolean;
     }): Promise<SubagentSpawnResult>;
     sendMessage(subagentId: string, message: string): Promise<SubagentSpawnResult>;
     stop(subagentId: string): Promise<void>;

package/dist/index.js

@@ -588,7 +588,8 @@ var defaultAgentDefinition = (opts = {}) => {
   const description = opts.description ?? DEFAULT_AGENT_DESCRIPTION;
   const modelProvider = opts.modelProvider ?? DEFAULT_MODEL_PROVIDER;
   const modelName = opts.modelName ?? DEFAULT_MODEL_NAME;
-  const temperature = opts.temperature ?? DEFAULT_TEMPERATURE;
+  const temperatureLine = opts.temperature !== void 0 ? `
+  temperature: ${opts.temperature}` : "";
   const maxSteps = opts.maxSteps ?? DEFAULT_MAX_STEPS;
   const timeout = opts.timeout ?? DEFAULT_TIMEOUT;
   return `---
@@ -597,8 +598,7 @@ id: ${id}
 description: ${description}
 model:
   provider: ${modelProvider}
-  name: ${modelName}
-  temperature: ${temperature}
+  name: ${modelName}${temperatureLine}
 limits:
   maxSteps: ${maxSteps}
   timeout: ${timeout}
@@ -8299,7 +8299,8 @@ var createSubagentTools = (manager) => [
         task: task.trim(),
         parentConversationId: conversationId,
         ownerId,
-        tenantId: context.tenantId
+        tenantId: context.tenantId,
+        suppressTelemetry: context.suppressTelemetry
       });
       return { subagentId, status: "running" };
     }
@@ -10831,7 +10832,7 @@ ${textContent}` };
           cachedCoreMessages = [...cachedCoreMessages, ...newCoreMessages];
           convertedUpTo = messages.length;
           const coreMessages = cachedCoreMessages;
-          const temperature = agent.frontmatter.model?.temperature ?? 0.2;
+          const temperature = agent.frontmatter.model?.temperature;
           const maxTokens = agent.frontmatter.model?.maxTokens;
           const cachedMessages = skipTailCache ? coreMessages : addPromptCacheBreakpoints(
             coreMessages,
@@ -10859,7 +10860,7 @@ ${textContent}` };
             ...useStaticCache ? {} : { system: systemPrompt },
             messages: messagesForStep,
             tools: toolsForStep,
-            temperature,
+            ...typeof temperature === "number" ? { temperature } : {},
             abortSignal: input.abortSignal,
             ...typeof maxTokens === "number" ? { maxTokens } : {},
             experimental_telemetry: {
@@ -11132,6 +11133,7 @@ ${textContent}` };
             abortSignal: input.abortSignal,
             conversationId: input.conversationId,
             tenantId: input.tenantId,
+            suppressTelemetry: input.suppressTelemetry,
             vfs: this.bashManager ? this.createVfsAccess(input.tenantId ?? "__default__") : void 0
           };
           const toolResultsForModel = [];
@@ -12903,7 +12905,9 @@ var AgentOrchestrator = class {
           __ownerId: ownerId
         }, conversation),
         messages: harnessMessages,
-        abortSignal: childAbortController.signal
+        abortSignal: childAbortController.signal,
+        // Inherit the parent run's telemetry choice (e.g. incognito).
+        suppressTelemetry: conversation.subagentMeta?.suppressTelemetry
       })) {
         if (event.type === "run:started") {
           latestRunId = event.runId;
@@ -13357,7 +13361,9 @@ ${resultBody}`,
           __ownerId: ownerId
         }, conversation),
         messages: continuationMessages,
-        abortSignal: childAbortController.signal
+        abortSignal: childAbortController.signal,
+        // Inherit the parent run's telemetry choice (e.g. incognito).
+        suppressTelemetry: conversation.subagentMeta?.suppressTelemetry
       })) {
         if (event.type === "run:started") {
           const active = this.activeConversationRuns.get(conversationId);
@@ -13527,7 +13533,7 @@ ${resultBody}`,
           opts.tenantId ?? null,
           {
             parentConversationId: opts.parentConversationId,
-            subagentMeta: { task: opts.task, status: "running" },
+            subagentMeta: { task: opts.task, status: "running", suppressTelemetry: opts.suppressTelemetry },
             messages: [{ role: "user", content: opts.task }]
           }
         );

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@poncho-ai/harness",
-  "version": "0.51.1",
+  "version": "0.52.1",
   "description": "Agent execution runtime - conversation loop, tool dispatch, streaming",
   "repository": {
     "type": "git",
@@ -34,7 +34,7 @@
     "mustache": "^4.2.0",
     "yaml": "^2.4.0",
     "zod": "^3.22.0",
-    "@poncho-ai/sdk": "1.13.0"
+    "@poncho-ai/sdk": "1.14.0"
   },
   "peerDependencies": {
     "esbuild": ">=0.17.0",

package/src/default-agent.ts

@@ -26,7 +26,12 @@ export interface DefaultAgentDefinitionOptions {
   modelProvider?: "anthropic" | "openai" | "openai-codex";
   /** Model name. Default: "claude-opus-4-5". */
   modelName?: string;
-  /** Sampling temperature. Default: 0.2. */
+  /**
+   * Sampling temperature. When unset, it is omitted from the generated
+   * frontmatter entirely and the harness sends no temperature (provider
+   * default). Newer models (Fable 5, Opus 4.7+) reject `temperature` — leave
+   * this unset for them.
+   */
   temperature?: number;
   /** Max tool-call steps per run. Default: 20. */
   maxSteps?: number;
@@ -55,7 +60,10 @@ export const defaultAgentDefinition = (
   const description = opts.description ?? DEFAULT_AGENT_DESCRIPTION;
   const modelProvider = opts.modelProvider ?? DEFAULT_MODEL_PROVIDER;
   const modelName = opts.modelName ?? DEFAULT_MODEL_NAME;
-  const temperature = opts.temperature ?? DEFAULT_TEMPERATURE;
+  // Opt-in: only emit a `temperature:` line when explicitly provided, so the
+  // harness sends no temperature otherwise (newer models reject it).
+  const temperatureLine =
+    opts.temperature !== undefined ? `\n  temperature: ${opts.temperature}` : "";
   const maxSteps = opts.maxSteps ?? DEFAULT_MAX_STEPS;
   const timeout = opts.timeout ?? DEFAULT_TIMEOUT;
 
@@ -65,8 +73,7 @@ id: ${id}
 description: ${description}
 model:
   provider: ${modelProvider}
-  name: ${modelName}
-  temperature: ${temperature}
+  name: ${modelName}${temperatureLine}
 limits:
   maxSteps: ${maxSteps}
   timeout: ${timeout}

package/src/harness.ts

@@ -2907,7 +2907,12 @@ Code is wrapped in an async IIFE — use \`return\` to return a value to the too
         convertedUpTo = messages.length;
         const coreMessages = cachedCoreMessages;
 
-        const temperature = agent.frontmatter.model?.temperature ?? 0.2;
+        // Only send temperature when the agent explicitly set one. Newer
+        // models (Fable 5, Opus 4.7+) removed sampling params entirely and
+        // return a 400 ("`temperature` is deprecated for this model") on any
+        // value — forcing a default here broke them. Treated like maxTokens
+        // below: omitted from the request when undefined.
+        const temperature = agent.frontmatter.model?.temperature;
         const maxTokens = agent.frontmatter.model?.maxTokens;
         // Place the tail breakpoint before any untruncated tool-result so
         // we cache only the stable prefix when prior-run tool results are
@@ -2971,7 +2976,7 @@ Code is wrapped in an async IIFE — use \`return\` to return a value to the too
           ...(useStaticCache ? {} : { system: systemPrompt }),
           messages: messagesForStep,
           tools: toolsForStep,
-          temperature,
+          ...(typeof temperature === "number" ? { temperature } : {}),
           abortSignal: input.abortSignal,
           ...(typeof maxTokens === "number" ? { maxTokens } : {}),
           experimental_telemetry: {
@@ -3285,6 +3290,7 @@ Code is wrapped in an async IIFE — use \`return\` to return a value to the too
         abortSignal: input.abortSignal,
         conversationId: input.conversationId,
         tenantId: input.tenantId,
+        suppressTelemetry: input.suppressTelemetry,
         vfs: this.bashManager
           ? this.createVfsAccess(input.tenantId ?? "__default__")
           : undefined,

package/src/orchestrator/orchestrator.ts

@@ -825,6 +825,8 @@ export class AgentOrchestrator {
         }, conversation),
         messages: harnessMessages,
         abortSignal: childAbortController.signal,
+        // Inherit the parent run's telemetry choice (e.g. incognito).
+        suppressTelemetry: conversation.subagentMeta?.suppressTelemetry,
       })) {
         if (event.type === "run:started") {
           latestRunId = event.runId;
@@ -1350,6 +1352,8 @@ export class AgentOrchestrator {
         }, conversation),
         messages: continuationMessages,
         abortSignal: childAbortController.signal,
+        // Inherit the parent run's telemetry choice (e.g. incognito).
+        suppressTelemetry: conversation.subagentMeta?.suppressTelemetry,
       })) {
         if (event.type === "run:started") {
           const active = this.activeConversationRuns.get(conversationId);
@@ -1530,7 +1534,7 @@ export class AgentOrchestrator {
           opts.tenantId ?? null,
           {
             parentConversationId: opts.parentConversationId,
-            subagentMeta: { task: opts.task, status: "running" },
+            subagentMeta: { task: opts.task, status: "running", suppressTelemetry: opts.suppressTelemetry },
             messages: [{ role: "user", content: opts.task }],
           },
         );

package/src/state.ts

@@ -75,6 +75,10 @@ export interface Conversation {
     status: "running" | "completed" | "error" | "stopped";
     result?: import("@poncho-ai/sdk").RunResult;
     error?: import("@poncho-ai/sdk").AgentFailure;
+    /** Inherited from the parent run at spawn time — when true, this
+     *  subagent's runs emit no telemetry (e.g. spawned from an incognito
+     *  turn). Read by the orchestrator's runSubagent / continuation. */
+    suppressTelemetry?: boolean;
   };
   channelMeta?: {
     platform: string;

package/src/subagent-manager.ts

@@ -37,6 +37,9 @@ export interface SubagentManager {
     parentConversationId: string;
     ownerId: string;
     tenantId?: string | null;
+    /** Inherit the parent run's telemetry choice — when true, the subagent
+     *  run (and its re-runs) emit no telemetry. */
+    suppressTelemetry?: boolean;
   }): Promise<SubagentSpawnResult>;
 
   sendMessage(subagentId: string, message: string): Promise<SubagentSpawnResult>;

package/src/subagent-tools.ts

@@ -45,6 +45,7 @@ export const createSubagentTools = (
         parentConversationId: conversationId,
         ownerId,
         tenantId: context.tenantId,
+        suppressTelemetry: context.suppressTelemetry,
       });
       return { subagentId, status: "running" };
     },