@tangle-network/agent-runtime 0.72.0 → 0.74.0

package/dist/{chunk-VLF5RHEQ.js → chunk-T2HVQVB4.js}

@@ -1,58 +1,3 @@
-// src/errors.ts
-import { AgentEvalError } from "@tangle-network/agent-eval";
-import {
-  AgentEvalError as AgentEvalError2,
-  ConfigError,
-  JudgeError,
-  NotFoundError,
-  ValidationError
-} from "@tangle-network/agent-eval";
-var SessionMismatchError = class extends AgentEvalError {
-  sessionBackend;
-  requestedBackend;
-  constructor(sessionBackend, requestedBackend, options) {
-    super(
-      "validation",
-      `Cannot resume ${sessionBackend} session with ${requestedBackend} backend`,
-      options
-    );
-    this.sessionBackend = sessionBackend;
-    this.requestedBackend = requestedBackend;
-  }
-};
-var BackendTransportError = class extends AgentEvalError {
-  backend;
-  status;
-  /**
-   * Truncated upstream response body (≤2 KiB) when available. Diagnostic
-   * only — surfaces in `backend_error.error.body` and `final.error.body`
-   * so operators can see "free_tier_limit", "invalid_api_key", etc. without
-   * cracking the log line open.
-   */
-  body;
-  constructor(backend, message, options) {
-    super("config", message, options);
-    this.backend = backend;
-    this.status = options?.status;
-    this.body = options?.body;
-  }
-};
-var RuntimeRunStateError = class extends AgentEvalError {
-  constructor(message, options) {
-    super("validation", message, options);
-  }
-};
-var PlannerError = class extends AgentEvalError {
-  constructor(message, options) {
-    super("validation", message, options);
-  }
-};
-var AnalystError = class extends AgentEvalError {
-  constructor(message, options) {
-    super("validation", message, options);
-  }
-};
-
 // src/runtime/sandbox-events.ts
 function extractLlmCallEvent(event, agentRunName) {
   if (!event || typeof event !== "object") return void 0;
@@ -127,17 +72,7 @@ function mapSandboxEvent(event, opts = {}) {
 }
 
 export {
-  SessionMismatchError,
-  BackendTransportError,
-  RuntimeRunStateError,
-  PlannerError,
-  AnalystError,
-  AgentEvalError2 as AgentEvalError,
-  ConfigError,
-  JudgeError,
-  NotFoundError,
-  ValidationError,
   extractLlmCallEvent,
   mapSandboxEvent
 };
-//# sourceMappingURL=chunk-VLF5RHEQ.js.map
+//# sourceMappingURL=chunk-T2HVQVB4.js.map

package/dist/chunk-T2HVQVB4.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/runtime/sandbox-events.ts"],"sourcesContent":["/**\n * Sandbox-event → runtime-event mapping.\n *\n * The sandbox SDK emits a polymorphic `SandboxEvent = { type, data, id? }`\n * whose `type` vocabulary is backend-determined (opencode, etc.) rather than\n * enumerated by the SDK. Two consumers project it:\n *   - the loop kernel's cost ledger (`extractLlmCallEvent`) — sums usage off\n *     every cost-bearing event, regardless of stream shape;\n *   - the `AgentRuntime.act` streaming contract (`mapSandboxEvent`) — projects\n *     incremental events to the `RuntimeStreamEvent` chat-UX vocabulary.\n *\n * Both live here so the empirically-observed `type` vocabulary has one home.\n */\n\nimport type { SandboxEvent } from '@tangle-network/sandbox'\nimport type { RuntimeStreamEvent } from '../types'\n\n/**\n * Extract a `RuntimeStreamEvent`-shaped `llm_call` from a sandbox event when\n * the event carries usage/cost data. Returns `undefined` for non-cost events\n * so the kernel can iterate the full stream without branching.\n *\n * Canonical cost-carrying types observed in the wild:\n *   - `llm_call` — `data: { model, tokensIn, tokensOut, costUsd, ... }`\n *   - `message.completed` / `result` — `data: { usage: { inputTokens,\n *      outputTokens, totalCostUsd? } }`\n *   - `cost.usage` / `usage` — same shape under a dedicated type\n *\n * Numeric coercion is strict: `Number.isFinite` gates every accumulator write\n * so a sentinel `NaN` from a misbehaving backend cannot poison the ledger.\n */\nexport function extractLlmCallEvent(\n  event: SandboxEvent,\n  agentRunName: string,\n): (RuntimeStreamEvent & { type: 'llm_call' }) | undefined {\n  if (!event || typeof event !== 'object') return undefined\n  const type = String(event.type ?? '')\n  const data =\n    event.data && typeof event.data === 'object'\n      ? (event.data as Record<string, unknown>)\n      : ({} as Record<string, unknown>)\n\n  if (type === 'llm_call' || type === 'cost.usage' || type === 'usage') {\n    return buildLlmCall(data, agentRunName)\n  }\n  if (type === 'message.completed' || type === 'result' || type === 'final') {\n    const usage = data.usage as Record<string, unknown> | undefined\n    if (!usage || typeof usage !== 'object') return undefined\n    return buildLlmCall({ ...usage, model: data.model ?? usage.model }, agentRunName)\n  }\n  // sandbox 0.4.0 terminal event: `data = { tokenUsage: { inputTokens, outputTokens,\n  // reasoningTokens, cacheReadInputTokens }, totalCostUsd }`. Usage lives under\n  // `tokenUsage` (not `usage`) and the cost is top-level — neither matched the\n  // branches above, so an in-process loopDispatch run reported {0,0} and the\n  // backend-integrity guard misread a real run as a stub. Reasoning tokens are\n  // billed output (reasoning models), so they fold into the output count.\n  if (type === 'done') {\n    const usage = data.tokenUsage as Record<string, unknown> | undefined\n    if (!usage || typeof usage !== 'object') return undefined\n    const out = pickFiniteNumber(usage, ['outputTokens', 'completion_tokens', 'tokensOut'])\n    const reasoning = pickFiniteNumber(usage, ['reasoningTokens'])\n    const mergedOut =\n      out !== undefined || reasoning !== undefined ? (out ?? 0) + (reasoning ?? 0) : undefined\n    return buildLlmCall(\n      {\n        inputTokens: usage.inputTokens,\n        outputTokens: mergedOut,\n        totalCostUsd: data.totalCostUsd,\n        model: data.model ?? usage.model,\n      },\n      agentRunName,\n    )\n  }\n  return undefined\n}\n\nfunction buildLlmCall(\n  data: Record<string, unknown>,\n  agentRunName: string,\n): (RuntimeStreamEvent & { type: 'llm_call' }) | undefined {\n  const tokensIn = pickFiniteNumber(data, ['tokensIn', 'inputTokens', 'prompt_tokens'])\n  const tokensOut = pickFiniteNumber(data, ['tokensOut', 'outputTokens', 'completion_tokens'])\n  const costUsd = pickFiniteNumber(data, ['costUsd', 'totalCostUsd', 'cost_usd', 'cost'])\n  if (tokensIn === undefined && tokensOut === undefined && costUsd === undefined) {\n    return undefined\n  }\n  const model = typeof data.model === 'string' && data.model.length > 0 ? data.model : agentRunName\n  const event: RuntimeStreamEvent & { type: 'llm_call' } = {\n    type: 'llm_call',\n    model,\n  }\n  if (tokensIn !== undefined) event.tokensIn = tokensIn\n  if (tokensOut !== undefined) event.tokensOut = tokensOut\n  if (costUsd !== undefined) event.costUsd = costUsd\n  return event\n}\n\nfunction pickFiniteNumber(data: Record<string, unknown>, keys: string[]): number | undefined {\n  for (const key of keys) {\n    const value = data[key]\n    if (typeof value === 'number' && Number.isFinite(value)) return value\n  }\n  return undefined\n}\n\n/**\n * Project one `SandboxEvent` onto the `RuntimeStreamEvent` chat-UX vocabulary,\n * for runtimes that bridge a sandbox `streamPrompt` into the\n * `AgentRuntime.act` streaming contract. Returns `undefined` for events that\n * have no faithful projection — the raw stream is preserved separately for the\n * `OutputAdapter`, so an unmapped event never loses data.\n *\n * Mapped (the task-optional incremental variants — no synthesized task\n * lifecycle, no guessed tool-part shapes):\n *   - `message.part.updated` text part → `text_delta`\n *   - `message.part.updated` reasoning/thinking part → `reasoning_delta`\n *   - cost-bearing events → `llm_call` (shared with the ledger extractor)\n *\n * The opencode backend emits incremental text as\n * `{ type: 'message.part.updated', data: { part: { type, text }, delta } }`;\n * `delta` is the increment, `part.text` the running accumulation.\n */\nexport function mapSandboxEvent(\n  event: SandboxEvent,\n  opts: { agentRunName?: string } = {},\n): RuntimeStreamEvent | undefined {\n  if (!event || typeof event !== 'object') return undefined\n  const type = String(event.type ?? '')\n  const data =\n    event.data && typeof event.data === 'object'\n      ? (event.data as Record<string, unknown>)\n      : ({} as Record<string, unknown>)\n\n  if (type === 'message.part.updated') {\n    const part =\n      data.part && typeof data.part === 'object' ? (data.part as Record<string, unknown>) : {}\n    const partType = String(part.type ?? '')\n    const delta = typeof data.delta === 'string' ? data.delta : undefined\n    const text = delta ?? (typeof part.text === 'string' ? part.text : undefined)\n    if (text === undefined) return undefined\n    if (partType === 'text') return { type: 'text_delta', text }\n    if (partType === 'reasoning' || partType === 'thinking')\n      return { type: 'reasoning_delta', text }\n    return undefined\n  }\n\n  return extractLlmCallEvent(event, opts.agentRunName ?? 'agent')\n}\n"],"mappings":";AA+BO,SAAS,oBACd,OACA,cACyD;AACzD,MAAI,CAAC,SAAS,OAAO,UAAU,SAAU,QAAO;AAChD,QAAM,OAAO,OAAO,MAAM,QAAQ,EAAE;AACpC,QAAM,OACJ,MAAM,QAAQ,OAAO,MAAM,SAAS,WAC/B,MAAM,OACN,CAAC;AAER,MAAI,SAAS,cAAc,SAAS,gBAAgB,SAAS,SAAS;AACpE,WAAO,aAAa,MAAM,YAAY;AAAA,EACxC;AACA,MAAI,SAAS,uBAAuB,SAAS,YAAY,SAAS,SAAS;AACzE,UAAM,QAAQ,KAAK;AACnB,QAAI,CAAC,SAAS,OAAO,UAAU,SAAU,QAAO;AAChD,WAAO,aAAa,EAAE,GAAG,OAAO,OAAO,KAAK,SAAS,MAAM,MAAM,GAAG,YAAY;AAAA,EAClF;AAOA,MAAI,SAAS,QAAQ;AACnB,UAAM,QAAQ,KAAK;AACnB,QAAI,CAAC,SAAS,OAAO,UAAU,SAAU,QAAO;AAChD,UAAM,MAAM,iBAAiB,OAAO,CAAC,gBAAgB,qBAAqB,WAAW,CAAC;AACtF,UAAM,YAAY,iBAAiB,OAAO,CAAC,iBAAiB,CAAC;AAC7D,UAAM,YACJ,QAAQ,UAAa,cAAc,UAAa,OAAO,MAAM,aAAa,KAAK;AACjF,WAAO;AAAA,MACL;AAAA,QACE,aAAa,MAAM;AAAA,QACnB,cAAc;AAAA,QACd,cAAc,KAAK;AAAA,QACnB,OAAO,KAAK,SAAS,MAAM;AAAA,MAC7B;AAAA,MACA;AAAA,IACF;AAAA,EACF;AACA,SAAO;AACT;AAEA,SAAS,aACP,MACA,cACyD;AACzD,QAAM,WAAW,iBAAiB,MAAM,CAAC,YAAY,eAAe,eAAe,CAAC;AACpF,QAAM,YAAY,iBAAiB,MAAM,CAAC,aAAa,gBAAgB,mBAAmB,CAAC;AAC3F,QAAM,UAAU,iBAAiB,MAAM,CAAC,WAAW,gBAAgB,YAAY,MAAM,CAAC;AACtF,MAAI,aAAa,UAAa,cAAc,UAAa,YAAY,QAAW;AAC9E,WAAO;AAAA,EACT;AACA,QAAM,QAAQ,OAAO,KAAK,UAAU,YAAY,KAAK,MAAM,SAAS,IAAI,KAAK,QAAQ;AACrF,QAAM,QAAmD;AAAA,IACvD,MAAM;AAAA,IACN;AAAA,EACF;AACA,MAAI,aAAa,OAAW,OAAM,WAAW;AAC7C,MAAI,cAAc,OAAW,OAAM,YAAY;AAC/C,MAAI,YAAY,OAAW,OAAM,UAAU;AAC3C,SAAO;AACT;AAEA,SAAS,iBAAiB,MAA+B,MAAoC;AAC3F,aAAW,OAAO,MAAM;AACtB,UAAM,QAAQ,KAAK,GAAG;AACtB,QAAI,OAAO,UAAU,YAAY,OAAO,SAAS,KAAK,EAAG,QAAO;AAAA,EAClE;AACA,SAAO;AACT;AAmBO,SAAS,gBACd,OACA,OAAkC,CAAC,GACH;AAChC,MAAI,CAAC,SAAS,OAAO,UAAU,SAAU,QAAO;AAChD,QAAM,OAAO,OAAO,MAAM,QAAQ,EAAE;AACpC,QAAM,OACJ,MAAM,QAAQ,OAAO,MAAM,SAAS,WAC/B,MAAM,OACN,CAAC;AAER,MAAI,SAAS,wBAAwB;AACnC,UAAM,OACJ,KAAK,QAAQ,OAAO,KAAK,SAAS,WAAY,KAAK,OAAmC,CAAC;AACzF,UAAM,WAAW,OAAO,KAAK,QAAQ,EAAE;AACvC,UAAM,QAAQ,OAAO,KAAK,UAAU,WAAW,KAAK,QAAQ;AAC5D,UAAM,OAAO,UAAU,OAAO,KAAK,SAAS,WAAW,KAAK,OAAO;AACnE,QAAI,SAAS,OAAW,QAAO;AAC/B,QAAI,aAAa,OAAQ,QAAO,EAAE,MAAM,cAAc,KAAK;AAC3D,QAAI,aAAa,eAAe,aAAa;AAC3C,aAAO,EAAE,MAAM,mBAAmB,KAAK;AACzC,WAAO;AAAA,EACT;AAEA,SAAO,oBAAoB,OAAO,KAAK,gBAAgB,OAAO;AAChE;","names":[]}

package/dist/{chunk-55WKGOGU.js → chunk-ZKMOIEOB.js}

@@ -1,127 +1,24 @@
 import {
   buildLoopSpanNodes
 } from "./chunk-VMNEQHJR.js";
+import {
+  extractLlmCallEvent
+} from "./chunk-T2HVQVB4.js";
+import {
+  harnessInvocation,
+  runLocalHarness
+} from "./chunk-O2UPHN7X.js";
 import {
   AgentEvalError,
   ConfigError,
   NotFoundError,
   RuntimeRunStateError,
-  ValidationError,
-  extractLlmCallEvent
-} from "./chunk-VLF5RHEQ.js";
+  ValidationError
+} from "./chunk-NBV35BR6.js";
 import {
   UI_LENSES
 } from "./chunk-WIR4HOOJ.js";
 
-// src/mcp/local-harness.ts
-import { spawn } from "child_process";
-var HARNESS_INVOCATIONS = {
-  claude: {
-    command: "claude",
-    buildArgs: (taskPrompt) => ["--headless", "-p", taskPrompt],
-    modelArgs: (model) => ["-m", model]
-  },
-  codex: {
-    command: "codex",
-    buildArgs: (taskPrompt) => ["run", taskPrompt],
-    modelArgs: (model) => ["-m", model]
-  },
-  opencode: {
-    command: "opencode",
-    buildArgs: (taskPrompt) => ["run", taskPrompt],
-    modelArgs: (model) => ["-m", model]
-  }
-};
-function harnessInvocation(harness, profile, taskPrompt) {
-  const invocation = HARNESS_INVOCATIONS[harness];
-  if (!invocation) {
-    throw new Error(`harnessInvocation: unknown harness ${String(harness)}`);
-  }
-  const systemPrompt = profile.prompt?.systemPrompt;
-  const composedPrompt = typeof systemPrompt === "string" && systemPrompt.trim().length > 0 ? `${systemPrompt}
-
-${taskPrompt}` : taskPrompt;
-  const args = invocation.buildArgs(composedPrompt);
-  const model = profile.model?.default;
-  if (typeof model === "string" && model.length > 0) {
-    args.push(...invocation.modelArgs(model));
-  }
-  return { command: invocation.command, args };
-}
-var DEFAULT_TIMEOUT_MS = 5 * 60 * 1e3;
-function runLocalHarness(options) {
-  const { harness, cwd, taskPrompt } = options;
-  const timeoutMs = options.timeoutMs ?? DEFAULT_TIMEOUT_MS;
-  const env = options.env ?? process.env;
-  const spawnImpl = options.spawn ?? spawn;
-  const invocation = HARNESS_INVOCATIONS[harness];
-  if (!invocation) {
-    return Promise.reject(new Error(`runLocalHarness: unknown harness ${String(harness)}`));
-  }
-  const startedAt = Date.now();
-  const command = options.invocation?.command ?? invocation.command;
-  const args = options.invocation ? [...options.invocation.args] : invocation.buildArgs(taskPrompt);
-  return new Promise((resolve, reject) => {
-    let child;
-    try {
-      child = spawnImpl(command, args, { cwd, env, stdio: "pipe" });
-    } catch (err) {
-      reject(err instanceof Error ? err : new Error(String(err)));
-      return;
-    }
-    child.stdin?.end();
-    let stdout = "";
-    let stderr = "";
-    let timedOut = false;
-    let settled = false;
-    const timer = timeoutMs > 0 ? setTimeout(() => {
-      timedOut = true;
-      if (!child.killed) child.kill("SIGTERM");
-    }, timeoutMs) : null;
-    if (timer && typeof timer.unref === "function") {
-      ;
-      timer.unref();
-    }
-    const onAbort = () => {
-      if (!child.killed) child.kill("SIGTERM");
-    };
-    if (options.signal) {
-      if (options.signal.aborted) onAbort();
-      else options.signal.addEventListener("abort", onAbort, { once: true });
-    }
-    child.stdout?.on("data", (chunk) => {
-      stdout += String(chunk);
-    });
-    child.stderr?.on("data", (chunk) => {
-      stderr += String(chunk);
-    });
-    const finalize3 = (result) => {
-      if (settled) return;
-      settled = true;
-      if (timer) clearTimeout(timer);
-      options.signal?.removeEventListener("abort", onAbort);
-      resolve(result);
-    };
-    child.on("error", (err) => {
-      if (settled) return;
-      settled = true;
-      if (timer) clearTimeout(timer);
-      options.signal?.removeEventListener("abort", onAbort);
-      reject(err);
-    });
-    child.on("close", (code, signal) => {
-      finalize3({
-        exitCode: code,
-        stdout,
-        stderr,
-        killedBySignal: signal,
-        durationMs: Date.now() - startedAt,
-        timedOut
-      });
-    });
-  });
-}
-
 // src/runtime/supervise/model-policy.ts
 function assertModelAllowed(model, allowed) {
   if (!allowed || model === void 0) return;
@@ -162,6 +59,9 @@ function throwAbort() {
 function throwIfAborted(signal) {
   if (signal?.aborted) throw abortError();
 }
+function isAbortError(err) {
+  return err instanceof Error && err.name === "AbortError";
+}
 function sleep(ms, signal) {
   return new Promise((resolve) => {
     if (signal?.aborted) {
@@ -564,6 +464,8 @@ function createSandboxLineage(client, capabilities, options = {}) {
     throw new ValidationError("createSandboxLineage: client.create is required");
   }
   const streaming = options.streaming ?? "sse";
+  const recordMount = options.recordMount ?? (() => {
+  });
   const forkConcurrency = Math.max(
     1,
     Math.floor(options.maxConcurrency ?? DEFAULT_FORK_CONCURRENCY)
@@ -573,7 +475,7 @@ function createSandboxLineage(client, capabilities, options = {}) {
     if (signal.aborted) throwAbort();
     const opts = buildBackendOptions(spec.profile, spec.sandboxOverrides);
     const box = await acquireSandbox(client, opts, { signal });
-    await spec.prepareBox?.(box, { signal });
+    await spec.prepareBox?.(box, { signal, recordMount });
     owned.push(box);
     return box;
   };
@@ -602,7 +504,7 @@ function createSandboxLineage(client, capabilities, options = {}) {
         if (checkpointId !== void 0) {
           const box2 = await forkFromCheckpoint(parent.box, checkpointId, signal);
           owned.push(box2);
-          await spec.prepareBox?.(box2, { signal });
+          await spec.prepareBox?.(box2, { signal, recordMount });
           const sessionId2 = mintSessionId();
           return {
             handle: { box: box2, sessionId: sessionId2 },
@@ -691,13 +593,17 @@ async function runLoop(options) {
   const loopStart = now();
   const driverName = options.driver.name ?? "driver";
   const iterations = [];
+  const mounts = [];
+  const recordMount = (entry) => {
+    mounts.push(entry);
+  };
   let round = 0;
   const ownedBoxes = [];
   const collectBox = options.onWorkerBox ? (box) => {
     ownedBoxes.push(box);
     options.onWorkerBox?.(box);
   } : void 0;
-  const lineageState = await setUpLineage(options, maxConcurrency);
+  const lineageState = await setUpLineage(options, maxConcurrency, recordMount);
   emitRunLoopHook(options, {
     target: "agent.run",
     phase: "before",
@@ -808,7 +714,8 @@ async function runLoop(options) {
         parentIndex,
         collectBox,
         lineagePlan,
-        lineageState
+        lineageState,
+        recordMount
       });
       if (controller.signal.aborted) throwAbort();
       emitRunLoopHook(options, {
@@ -835,11 +742,19 @@ async function runLoop(options) {
         payload: { decision: stringifySafe(decision), historyLength: iterations.length }
       });
       if (isTerminalDecision(decision)) {
-        return await finalizeAndEmitEnded(options, decision, iterations, loopStart, now, runId);
+        return await finalizeAndEmitEnded(
+          options,
+          decision,
+          iterations,
+          loopStart,
+          now,
+          runId,
+          mounts
+        );
       }
       if (lineageState) await pruneLineage(lineageState, iterations);
     }
-    return await decideAndFinalize(options, iterations, loopStart, now, runId);
+    return await decideAndFinalize(options, iterations, loopStart, now, runId, mounts);
   } finally {
     if (options.ctx.signal) options.ctx.signal.removeEventListener("abort", onOuterAbort);
     await Promise.allSettled(
@@ -849,7 +764,7 @@ async function runLoop(options) {
     if (lineageState) await lineageState.lineage.teardown();
   }
 }
-async function setUpLineage(options, maxConcurrency) {
+async function setUpLineage(options, maxConcurrency, recordMount) {
   const lineageOpts = options.lineage;
   if (!lineageOpts || !lineageOpts.sessionContinuity && !lineageOpts.forkFanout) return void 0;
   if (options.onWorkerBox) {
@@ -861,7 +776,8 @@ async function setUpLineage(options, maxConcurrency) {
   return {
     lineage: createSandboxLineage(options.ctx.sandboxClient, capabilities, {
       maxConcurrency,
-      streaming: lineageOpts.streaming
+      streaming: lineageOpts.streaming,
+      recordMount
     }),
     options: lineageOpts,
     handles: /* @__PURE__ */ new Map(),
@@ -991,7 +907,7 @@ async function executeIteration(args) {
       args.lineageState?.handles.set(args.item.index, acquired.handle);
       stream = acquired.events;
     } else {
-      box = await createSandboxForSpec(args.ctx.sandboxClient, spec, args.signal);
+      box = await createSandboxForSpec(args.ctx.sandboxClient, spec, args.signal, args.recordMount);
       const prompt = spec.taskToPrompt(args.item.task);
       stream = args.streaming === "poll" ? promptEvents("poll", box, prompt, `${args.runId}-i${args.item.index}`, args.signal) : box.streamPrompt(prompt, { signal: args.signal });
     }
@@ -1060,13 +976,13 @@ async function executeIteration(args) {
       await destroySandboxSafe(box, args.ctx.traceEmitter, args.runId, args.now);
     }
   }
-  if (isAbortError(slot.error) || args.signal.aborted) {
+  if (isAbortError2(slot.error) || args.signal.aborted) {
     if (slot.error) throw slot.error;
     throwAbort();
   }
   if (slot.error instanceof ValidationError) throw slot.error;
 }
-function isAbortError(err) {
+function isAbortError2(err) {
   return err instanceof Error && err.name === "AbortError";
 }
 var TEARDOWN_TIMEOUT_MS2 = 15e3;
@@ -1120,15 +1036,35 @@ function readSandboxId(box) {
   const raw = box.id;
   return typeof raw === "string" && raw.length > 0 ? raw : void 0;
 }
-async function createSandboxForSpec(client, spec, signal) {
+async function createSandboxForSpec(client, spec, signal, recordMount) {
   const opts = buildBackendOptions(spec.profile, spec.sandboxOverrides);
   if (signal.aborted) throwAbort();
   const box = await acquireSandbox(client, opts, { signal });
-  await spec.prepareBox?.(box, { signal });
+  await invokePrepareBox(spec, box, signal, recordMount);
   return box;
 }
+async function invokePrepareBox(spec, box, signal, recordMount) {
+  if (!spec.prepareBox) return;
+  await spec.prepareBox(box, { signal, recordMount: recordMount ?? noopMountRecorder });
+}
+var noopMountRecorder = () => {
+};
 function finalize(args) {
-  const winner = args.options.selectWinner ? args.options.selectWinner(args.iterations) : args.options.driver.selectWinner?.(args.iterations) ?? defaultSelectWinner(args.iterations);
+  let selector;
+  let winner;
+  if (args.options.selectWinner) {
+    selector = "caller";
+    winner = args.options.selectWinner(args.iterations);
+  } else {
+    const authored = args.options.driver.selectWinner?.(args.iterations);
+    if (authored) {
+      selector = "driver";
+      winner = authored;
+    } else {
+      selector = "default";
+      winner = defaultSelectWinner(args.iterations);
+    }
+  }
   const costUsd = args.iterations.reduce((sum, iter) => sum + (iter.costUsd || 0), 0);
   const tokenUsage = args.iterations.reduce((acc, iter) => {
     addTokenUsage(acc, iter.tokenUsage);
@@ -1140,11 +1076,36 @@ function finalize(args) {
     winner,
     durationMs: args.now() - args.startMs,
     costUsd,
-    tokenUsage
+    tokenUsage,
+    provenance: {
+      mounts: args.mounts,
+      selectionReceipts: buildSelectionReceipts(args.iterations, winner, selector)
+    }
   };
   return result;
 }
-async function decideAndFinalize(options, iterations, startMs, now, runId) {
+function buildSelectionReceipts(iterations, winner, selector) {
+  const receipts = [];
+  for (const iter of iterations) {
+    if (iter.output === void 0 || iter.error) continue;
+    const selected = winner?.iterationIndex === iter.index;
+    const receipt = {
+      candidateIndex: iter.index,
+      selected,
+      selector
+    };
+    if (iter.verdict?.score !== void 0) receipt.score = iter.verdict.score;
+    if (selector === "default") receipt.reason = defaultSelectorReason(iter, selected);
+    receipts.push(receipt);
+  }
+  return receipts;
+}
+function defaultSelectorReason(iter, selected) {
+  const valid = iter.verdict?.valid === true;
+  if (selected) return valid ? "best valid score" : "best non-errored score (no valid candidate)";
+  return valid ? "valid but not top score" : "not selected";
+}
+async function decideAndFinalize(options, iterations, startMs, now, runId, mounts) {
   emitRunLoopHook(options, {
     target: "agent.decision",
     phase: "before",
@@ -1166,10 +1127,10 @@ async function decideAndFinalize(options, iterations, startMs, now, runId) {
     timestamp: now(),
     payload: { decision: stringifySafe(decision), historyLength: iterations.length }
   });
-  return finalizeAndEmitEnded(options, decision, iterations, startMs, now, runId);
+  return finalizeAndEmitEnded(options, decision, iterations, startMs, now, runId, mounts);
 }
-async function finalizeAndEmitEnded(options, decision, iterations, startMs, now, runId) {
-  const result = finalize({ options, decision, iterations, startMs, now, runId });
+async function finalizeAndEmitEnded(options, decision, iterations, startMs, now, runId, mounts) {
+  const result = finalize({ options, decision, iterations, startMs, now, runId, mounts });
   emitRunLoopHook(options, {
     target: "agent.run",
     phase: "after",
@@ -1293,7 +1254,7 @@ function createScope(args) {
   let cursorSeq = 0;
   let meterSeq = 0;
   const now = args.now ?? Date.now;
-  function spawn5(agent, task, opts) {
+  function spawn4(agent, task, opts) {
     if (args.maxDepth !== void 0 && args.depth >= args.maxDepth) {
       return { ok: false, reason: "depth-exceeded" };
     }
@@ -1444,7 +1405,7 @@ function createScope(args) {
     );
   }
   return {
-    spawn: spawn5,
+    spawn: spawn4,
     next,
     send,
     signal: args.signal,
@@ -1605,7 +1566,7 @@ async function runChild(live, executor, childAbort, task, opts, pool, ticket, bl
   } catch (err) {
     reconcileOnce(live.spent);
     await teardownSafe(executor, "brutalKill");
-    const aborted = childAbort.signal.aborted || isAbortError2(err);
+    const aborted = childAbort.signal.aborted || isAbortError3(err);
     return downRecord(errMessage(err), aborted || isInfraError(err), executor.metered?.());
   }
 }
@@ -1707,7 +1668,7 @@ function isAgentSpec(value) {
   const v = value;
   return "profile" in v && "harness" in v;
 }
-function isAbortError2(err) {
+function isAbortError3(err) {
   return typeof err === "object" && err !== null && "name" in err && err.name === "AbortError";
 }
 function isInfraError(err) {
@@ -1916,11 +1877,11 @@ ${lines.join("\n")}`;
 }
 
 // src/mcp/worktree.ts
-import { spawn as spawn2 } from "child_process";
+import { spawn } from "child_process";
 async function runGitAsync(args, cwd, runner) {
   if (runner) return runner(args, { cwd });
   return new Promise((resolve, reject) => {
-    const proc = spawn2("git", args, { cwd, stdio: "pipe" });
+    const proc = spawn("git", args, { cwd, stdio: "pipe" });
     let stdout = "";
     let stderr = "";
     proc.stdout?.on("data", (c) => {
@@ -2005,7 +1966,7 @@ async function removeWorktree(options) {
 import { randomUUID } from "crypto";
 
 // src/mcp/worktree-harness.ts
-import { spawn as spawn3 } from "child_process";
+import { spawn as spawn2 } from "child_process";
 var defaultCheckOutputCap = 16e3;
 async function runWorktreeHarness(opts) {
   const runHarness = opts.runHarness ?? runLocalHarness;
@@ -2090,7 +2051,7 @@ async function runChecks(opts) {
 }
 function defaultRunCommand(opts) {
   return new Promise((resolve, reject) => {
-    const child = spawn3("/bin/sh", ["-c", opts.command], {
+    const child = spawn2("/bin/sh", ["-c", opts.command], {
       cwd: opts.cwd,
       stdio: ["ignore", "pipe", "pipe"]
     });
@@ -2126,12 +2087,12 @@ function createWorktreeCliExecutor(options) {
   }
   const runId = options.runId ?? randomUUID();
   const controller = new AbortController();
+  const budgetExempt = options.budgetExempt ?? true;
   let run;
   let artifact;
   return {
     runtime: "cli",
-    // A harness CLI cannot account tokens — exclude it from the conserved pool + equal-k.
-    budgetExempt: true,
+    budgetExempt,
     async execute(_task, signal) {
       const linked = linkSignals(signal, controller.signal);
       const started = Date.now();
@@ -2153,7 +2114,9 @@ function createWorktreeCliExecutor(options) {
       });
       const spent = {
         iterations: 1,
-        // budgetExempt: spend is recorded zero (not metered), never a fabricated cost.
+        // The worktree-harness core surfaces no token/usd usage, so tokens/usd are a genuine zero
+        // (NOT a fabricated cost). When budgetExempt is true the pool ignores this spend entirely;
+        // when explicitly false the scope debits exactly this captured spend — the real iteration.
         tokens: zeroTokenUsage(),
         usd: 0,
         ms: Date.now() - started
@@ -2194,7 +2157,7 @@ function linkSignals(a, b) {
 }
 
 // src/runtime/supervise/runtime.ts
-import { spawn as spawn4 } from "child_process";
+import { spawn as spawn3 } from "child_process";
 import { randomUUID as randomUUID2 } from "crypto";
 import { estimateCost as estimateCost2, isModelPriced as isModelPriced2 } from "@tangle-network/agent-eval";
 var routerSeamKey = "router";
@@ -2585,7 +2548,7 @@ var cliExecutor = (_spec, ctx) => {
 };
 async function* streamCliLeaf(args) {
   const prompt = taskToPrompt(args.task);
-  const proc = spawn4(args.seam.bin, args.seam.args ?? [], {
+  const proc = spawn3(args.seam.bin, args.seam.args ?? [], {
     ...args.seam.cwd ? { cwd: args.seam.cwd } : {},
     env: { ...process.env, ...args.seam.env ?? {} },
     stdio: ["pipe", "pipe", "pipe"]
@@ -3769,10 +3732,14 @@ function createCoordinationTools(opts) {
       return q.status !== "answered" && q.status !== "deferred";
     });
   };
+  const maxLiveWorkers = opts.maxLiveWorkers;
+  const liveWorkerCount = () => opts.scope.view.nodes.filter(
+    (n) => n.status !== "done" && n.status !== "failed" && n.status !== "cancelled"
+  ).length;
   const tools = [
     {
       name: "spawn_agent",
-      description: "Start a worker the driver will drive. `profile` is the worker or another driver; `task` is what it should do. Reserves budget from the conserved pool and fails closed. Pass an optional `budget` (per-field) to give a hard sub-task more than the default \u2014 it merges over the per-worker default; the conserved pool is still the hard fence.",
+      description: 'Start a worker the driver will drive. `profile` is the worker or another driver; `task` is what it should do. Reserves budget from the conserved pool and fails closed. Pass an optional `budget` (per-field) to give a hard sub-task more than the default \u2014 it merges over the per-worker default; the conserved pool is still the hard fence. When a max-live-workers cap is set it also fails closed (`error: "max-live-workers"`) while that many workers are still in flight \u2014 settle or steer one before spawning another.',
       inputSchema: {
         type: "object",
         properties: {
@@ -3794,6 +3761,8 @@ function createCoordinationTools(opts) {
       },
       handler: (raw) => {
         const a = obj(raw);
+        if (maxLiveWorkers !== void 0 && maxLiveWorkers > 0 && liveWorkerCount() >= maxLiveWorkers)
+          return Promise.resolve({ error: "max-live-workers" });
         const agent = opts.makeWorkerAgent(a.profile);
         const budget = a.budget === void 0 ? opts.perWorker : mergeBudget(opts.perWorker, a.budget);
         const res = opts.scope.spawn(agent, a.task, {
@@ -4073,7 +4042,8 @@ function driverAgent(opts) {
         scope,
         blobs: opts.blobs,
         makeWorkerAgent: opts.makeWorkerAgent,
-        perWorker: opts.perWorker
+        perWorker: opts.perWorker,
+        ...opts.maxLiveWorkers !== void 0 ? { maxLiveWorkers: opts.maxLiveWorkers } : {}
       });
       const byName = new Map(coord.tools.map((t) => [t.name, t]));
       const toolSpecs = [
@@ -5157,6 +5127,7 @@ function supervise(profile, task, opts) {
     blobs,
     makeWorkerAgent,
     perWorker,
+    ...opts.maxLiveWorkers !== void 0 ? { maxLiveWorkers: opts.maxLiveWorkers } : {},
     ...opts.router ? { router: opts.router } : {},
     ...opts.brain ? { brain: opts.brain } : {},
     ...opts.driveHarness ? { driveHarness: opts.driveHarness } : {},
@@ -6036,6 +6007,7 @@ async function serveCoordinationMcp(opts) {
     blobs: opts.blobs,
     makeWorkerAgent: opts.makeWorkerAgent,
     perWorker: opts.perWorker,
+    ...opts.maxLiveWorkers !== void 0 ? { maxLiveWorkers: opts.maxLiveWorkers } : {},
     ...opts.analysts ? { analysts: opts.analysts } : {},
     ...opts.analyzeOnSettle ? { analyzeOnSettle: opts.analyzeOnSettle } : {},
     ...opts.onEvent ? { onEvent: opts.onEvent } : {},
@@ -6112,6 +6084,7 @@ function supervisorAgent(profile, deps) {
       makeWorkerAgent: deps.makeWorkerAgent,
       perWorker: deps.perWorker,
       systemPrompt,
+      ...deps.maxLiveWorkers !== void 0 ? { maxLiveWorkers: deps.maxLiveWorkers } : {},
       ...deps.extraTools ? { extraTools: deps.extraTools } : {},
       ...deps.executeExtraTool ? { executeExtraTool: deps.executeExtraTool } : {},
       ...deps.maxTurns !== void 0 ? { maxTurns: deps.maxTurns } : {}
@@ -6130,7 +6103,8 @@ function supervisorAgent(profile, deps) {
         scope,
         blobs: deps.blobs,
         makeWorkerAgent: deps.makeWorkerAgent,
-        perWorker: deps.perWorker
+        perWorker: deps.perWorker,
+        ...deps.maxLiveWorkers !== void 0 ? { maxLiveWorkers: deps.maxLiveWorkers } : {}
       });
       try {
         await driveHarness({ profile, task, scope, coordinationMcpUrl: mcp.url });
@@ -6151,12 +6125,12 @@ function routerBrainFromProfile(profile, deps) {
 }
 
 export {
-  runLocalHarness,
   assertModelAllowed,
   deleteBoxSafe,
   randomSuffix,
   throwAbort,
   throwIfAborted,
+  isAbortError,
   sleep,
   stringifySafe,
   zeroTokenUsage,
@@ -6253,4 +6227,4 @@ export {
   createInProcessTransport,
   serveCoordinationMcp
 };
-//# sourceMappingURL=chunk-55WKGOGU.js.map
+//# sourceMappingURL=chunk-ZKMOIEOB.js.map