@tangle-network/agent-runtime 0.343.4 → 0.348.1

package/dist/{delegate-CjdtLf1v.js → delegate-Cp5-PsaX.js}

@@ -1,5 +1,5 @@
 import { i as ConfigError } from "./errors-DodWX-cb.js";
-import { t as supervise } from "./supervise-BuVVC7Bl.js";
+import { t as supervise } from "./supervise-BUOeThuB.js";
 import { agentProfileSchema } from "@tangle-network/agent-interface";
 import { computeFindingId, makeFinding } from "@tangle-network/agent-eval";
 //#region src/runtime/supervise/authoring.ts
@@ -207,4 +207,4 @@ async function delegate(intent, opts) {
 //#endregion
 export { defaultProfileRichnessThresholds as a, assessAuthoredProfile as i, delegate as n, profileRichnessFinding as o, asAuthoredProfile as r, supervisorInstructions as s, defaultDelegateBudget as t };
 
-//# sourceMappingURL=delegate-CjdtLf1v.js.map
+//# sourceMappingURL=delegate-Cp5-PsaX.js.map

package/dist/{delegate-CjdtLf1v.js.map → delegate-Cp5-PsaX.js.map}

@@ -1 +1 @@
-{"version":3,"file":"delegate-CjdtLf1v.js","names":[],"sources":["../src/runtime/supervise/authoring.ts","../src/runtime/supervise/delegate.ts"],"sourcesContent":["/**\n *\n * The supervisor's intelligence is AUTHORING the agents it spawns — not pressing buttons.\n *\n * Every agent here is three things: instructions (system prompt), tools, and a model — its\n * `AgentProfile`. The supervisor's job is to WRITE those profiles: read the task, decompose it,\n * and for each sub-task author a tailored worker recipe. `supervisorInstructions` is the how-to the\n * supervisor reads; canonical Runtime executors materialize the resulting profile.\n *\n * The skill is the single OPTIMIZABLE surface: edit it → the supervisor designs better agents.\n * That is the self-improvement lever (the prompt/skill lever), not the execution plumbing.\n *\n * @experimental\n */\n\nimport { type AnalystFinding, computeFindingId, makeFinding } from '@tangle-network/agent-eval'\nimport {\n  type AgentProfile,\n  type AgentProfilePrompt,\n  agentProfileSchema,\n} from '@tangle-network/agent-interface'\n\n/** What the supervisor AUTHORS per sub-task: one complete canonical profile whose name and\n *  task-specific system prompt are present. Every other `AgentProfile` axis is preserved exactly. */\nexport type AuthoredProfile = AgentProfile & {\n  readonly name: string\n  readonly prompt: AgentProfilePrompt & { readonly systemPrompt: string }\n}\n\n/** Narrow an untyped `spawn_worker` profile argument to an `AuthoredProfile`, or null if the\n *  supervisor failed to author one (empty/placeholder profile — a skill violation worth catching). */\nexport function asAuthoredProfile(raw: unknown): AuthoredProfile | null {\n  const parsed = agentProfileSchema.safeParse(raw)\n  if (!parsed.success) return null\n  const systemPrompt = parsed.data.prompt?.systemPrompt\n  if (typeof systemPrompt !== 'string' || systemPrompt.trim().length === 0) return null\n  return {\n    ...parsed.data,\n    name:\n      typeof parsed.data.name === 'string' && parsed.data.name.length > 0\n        ? parsed.data.name\n        : 'worker',\n    prompt: { ...parsed.data.prompt, systemPrompt },\n  }\n}\n\n/** The supervisor skill: an explicit profile-authoring instruction, never an implicit Runtime\n * policy. Editing this text changes how a profile designs the descendants it spawns. */\nexport function supervisorInstructions(opts?: { goal?: string }): string {\n  return [\n    'Your delegation craft is AUTHORING: a spawned worker is exactly as good as the profile you write.',\n    '',\n    'For the task you are given:',\n    '1. DECOMPOSE it into the smallest set of sub-tasks a single focused worker can each deliver.',\n    '2. For EACH sub-task, AUTHOR a worker by calling spawn_worker with a COMPLETE `profile`:',\n    '   • name and description: who this specialist is and why it exists.',\n    '   • prompt.systemPrompt: rich instructions for THIS sub-task — exact output, process, evidence, and what \"done\" means.',\n    '   • model.default, model.reasoningEffort, and harness: choose the execution system deliberately when the task benefits from it.',\n    '   • tools, mcp, resources.skills/files/instructions, hooks, subagents, permissions, and modes: grant or attach every capability the worker needs; omit an axis only when it is intentionally unnecessary.',\n    '   • tools.agent_runtime_coordination_spawn_worker: true ONLY when this child should author and drive descendants. Add only the other agent_runtime_coordination_<verb> tools it will call, such as observe_agent or steer_agent.',\n    '   • A child with spawn_worker MUST carry this complete profile-authoring instruction as an immutable resources.skills entry with resources.failOnError: true, so it can author its own descendants from the same contract.',\n    '   • metadata may describe the work, but it never grants recursion or selects a Runtime execution path.',\n    '   NEVER spawn a worker with an empty profile. The quality of the worker IS the quality of the profile you write.',\n    '3. End your turn while your workers run: Runtime wakes you with each settled worker. A settlement says valid:true only if the deployable check passed.',\n    '4. If a worker did NOT deliver, AUTHOR A NEW profile whose prompt.systemPrompt names the SPECIFIC failure and how to fix it — never just retry the same profile.',\n    \"5. read_journal to re-read YOUR OWN record before you decide the next move: every spawn you made, every settle, every question and answer, every steer, every analyst finding — oldest first, this node only, including what you did before a restart. Use it to see what you already tried instead of trying it again. It is paged: pass the returned nextRow as the next call's sinceRow, narrow with kinds, and raise limit/maxBytes only as far as you will actually read. A truncated:true page means a bound cut it short — keep paging before you conclude you have read everything.\",\n    '6. AUTHOR YOUR OWN LENS when the questions you can already ask of a settled trace do not cover the failure you are chasing: define_analyst takes an id, a description, an area, the question in your own words, the instructions for answering it with trace evidence, and the smallest toolGroup that can answer it (model is the seat it runs on; omit it for the run default). It is DATA, never code. Then run_analyst it on any settled worker like a lens the run shipped with, and read the finding. list_analysts shows what you have. Define a lens when you need a different question asked — not a second copy of a question already on the menu.',\n    '7. EVERY refusal you get back carries a `reason` naming the exact unmet condition. Read it and change that condition — a spawn refused for budget-exhausted needs a smaller budget or the refund of a settled worker, an invalid-profile needs the named field fixed, a submit_result refused because the check THREW is a broken check to report, not a result to resubmit. Never repeat a call that was refused without changing what it was refused for.',\n    '8. ask_parent ONLY when you genuinely cannot decide, and then READ ITS OUTCOME. \"queued-for-parent\" means an inbox above you now holds the question. \"no-parent\" means no inbox above you is configured to receive it: the question is still on the run record for anyone watching, but nothing will route an answer back to you, so do not block. Decide it with answer_question, or answer_question with deferReason to record that it stays open, and carry on — a blocking question left undecided also refuses your stop.',\n    '9. Once the work is delivered, end your turn with nothing left running.',\n    ...(opts?.goal ? ['', `The goal: ${opts.goal}`] : []),\n  ].join('\\n')\n}\n\n// ── Profile-richness gate ────────────────────────────────────────────────────\n//\n// The supervisor's product is the worker PROFILE it authors. The failure mode the existing\n// gates miss: `asAuthoredProfile` / `local-harness` only reject a FULLY EMPTY system prompt —\n// a two-sentence stub passes. `assessAuthoredProfile` OBSERVES the authored artifact (it reads\n// no judge verdict, so it steers cleanly past `assertTraceDerivedFindings`) and flags THIN:\n// a short/few-line system prompt, OR no tools, OR no skills, OR no MCP when the task needs one.\n// It emits a real `AnalystFinding` so it rides the SAME coordination bus the driver pulls via\n// receives in its next wake — the supervisor can self-correct and re-author richer.\n\n/** Thresholds below which a system prompt is treated as a thin stub. Tunable per call. */\nexport interface ProfileRichnessThresholds {\n  /** A prompt shorter than this many characters is thin (default 600). */\n  readonly minSystemPromptChars: number\n  /** A prompt with fewer than this many non-blank lines is thin (default 6). */\n  readonly minSystemPromptLines: number\n}\n\n/** Default thresholds for `ProfileRichnessThresholds` — 600 chars / 6 lines minimum system prompt. */\nexport const defaultProfileRichnessThresholds: ProfileRichnessThresholds = {\n  minSystemPromptChars: 600,\n  minSystemPromptLines: 6,\n}\n\n/** Per-field verdict on one authored profile — the raw material the bench renders + scores. */\nexport interface ProfileRichness {\n  readonly name: string\n  /** The resolved system prompt (canonical `prompt.systemPrompt`, the sandbox `prompt.system`\n   *  convention, or a bare-string prompt — whichever the author used). */\n  readonly systemPrompt: string\n  readonly systemPromptChars: number\n  readonly systemPromptLines: number\n  readonly sentenceCount: number\n  readonly hasDescription: boolean\n  readonly hasTools: boolean\n  readonly hasSkills: boolean\n  readonly hasMcp: boolean\n  readonly hasSubagents: boolean\n  /** 0..1 — fraction of richness signals present (prompt-depth + the four levers). */\n  readonly richness: number\n  /** True when the supervisor authored a stub instead of a real profile. */\n  readonly thin: boolean\n  /** The specific reasons it is thin (empty when rich) — used in the finding's action. */\n  readonly reasons: string[]\n}\n\n/** Read the system prompt from any authored shape: canonical `prompt.systemPrompt`, the sandbox\n *  `prompt.system` convention, or a bare-string `prompt`. */\nfunction resolveSystemPrompt(profile: AgentProfile): string {\n  const pr = (profile as { prompt?: unknown }).prompt\n  if (typeof pr === 'string') return pr\n  if (pr && typeof pr === 'object') {\n    const o = pr as { systemPrompt?: unknown; system?: unknown }\n    if (typeof o.systemPrompt === 'string') return o.systemPrompt\n    if (typeof o.system === 'string') return o.system\n  }\n  return ''\n}\n\n/** OBSERVE one authored `AgentProfile` and score its richness (no judge verdict is read). The task\n *  context (`needsMcp`) lets a domain say \"this work needs a data/tool MCP\" so a missing MCP counts. */\nexport function assessAuthoredProfile(\n  profile: AgentProfile,\n  opts?: { needsMcp?: boolean; thresholds?: Partial<ProfileRichnessThresholds> },\n): ProfileRichness {\n  const th = { ...defaultProfileRichnessThresholds, ...(opts?.thresholds ?? {}) }\n  const systemPrompt = resolveSystemPrompt(profile)\n  const trimmed = systemPrompt.trim()\n  const systemPromptChars = trimmed.length\n  const systemPromptLines = trimmed\n    ? trimmed.split('\\n').filter((l) => l.trim().length > 0).length\n    : 0\n  const sentenceCount = trimmed\n    ? (trimmed.match(/[.!?](\\s|$)/g) ?? []).length || (trimmed ? 1 : 0)\n    : 0\n  const hasDescription =\n    typeof profile.description === 'string' && profile.description.trim().length > 0\n  const tools = (profile as { tools?: Record<string, unknown> }).tools\n  const hasTools = !!tools && Object.keys(tools).length > 0\n  const skills = (profile.resources as { skills?: unknown[] } | undefined)?.skills\n  const hasSkills = Array.isArray(skills) && skills.length > 0\n  const mcp = (profile as { mcp?: Record<string, unknown> }).mcp\n  const hasMcp = !!mcp && Object.keys(mcp).length > 0\n  const subagents = (profile as { subagents?: Record<string, unknown> }).subagents\n  const hasSubagents = !!subagents && Object.keys(subagents).length > 0\n\n  const reasons: string[] = []\n  const promptThin =\n    systemPromptChars < th.minSystemPromptChars || systemPromptLines < th.minSystemPromptLines\n  if (promptThin)\n    reasons.push(\n      `system prompt is thin (${systemPromptChars} chars, ${systemPromptLines} lines; need ≥${th.minSystemPromptChars} chars and ≥${th.minSystemPromptLines} lines)`,\n    )\n  if (!hasTools)\n    reasons.push('no tools granted (a worker can only act through the tools you grant it)')\n  if (!hasSkills) reasons.push('no skills attached (no reusable how-to notes injected)')\n  if (opts?.needsMcp && !hasMcp) reasons.push('no MCP server, but the task needs data/tool access')\n\n  // Richness = fraction of signals present. Prompt-depth is one signal; the four levers are the rest.\n  const signals = [!promptThin, hasTools, hasSkills, hasDescription, opts?.needsMcp ? hasMcp : true]\n  const richness = signals.filter(Boolean).length / signals.length\n  // THIN ⟺ the prompt is a stub OR the worker has no levers at all (no tools AND no skills AND no mcp).\n  const thin = promptThin || (!hasTools && !hasSkills && !hasMcp)\n\n  return {\n    name: profile.name ?? 'worker',\n    systemPrompt,\n    systemPromptChars,\n    systemPromptLines,\n    sentenceCount,\n    hasDescription,\n    hasTools,\n    hasSkills,\n    hasMcp,\n    hasSubagents,\n    richness,\n    thin,\n    reasons,\n  }\n}\n\n/** Turn a {@link ProfileRichness} verdict into a bus-routable `AnalystFinding` (area `profile-quality`).\n *  Severity scales with thinness; the recommended action names the MISSING lever so the supervisor can\n *  re-author. `subject` = the worker name so per-worker findings diff cleanly across re-authors. */\nexport function profileRichnessFinding(\n  richness: ProfileRichness,\n  opts?: { analystId?: string; runId?: string },\n): AnalystFinding {\n  const analyst_id = opts?.analystId ?? 'profile-richness'\n  const subject = richness.name\n  const claim = richness.thin\n    ? `Worker \"${richness.name}\" was authored as a THIN profile: ${richness.reasons.join('; ')}.`\n    : `Worker \"${richness.name}\" was authored as a rich profile (richness ${(richness.richness * 100).toFixed(0)}%).`\n  const severity: AnalystFinding['severity'] = richness.thin\n    ? richness.richness < 0.25\n      ? 'high'\n      : 'medium'\n    : 'info'\n  return makeFinding({\n    analyst_id,\n    severity,\n    area: 'profile-quality',\n    claim,\n    subject,\n    confidence: 0.9,\n    evidence_refs: [\n      {\n        kind: 'metric',\n        uri: `profile:${subject}`,\n        excerpt: `chars=${richness.systemPromptChars} lines=${richness.systemPromptLines} tools=${richness.hasTools} skills=${richness.hasSkills} mcp=${richness.hasMcp} richness=${richness.richness.toFixed(2)}`,\n      },\n    ],\n    ...(richness.thin\n      ? { recommended_action: `Re-author \"${richness.name}\" with: ${richness.reasons.join('; ')}.` }\n      : {}),\n    id_basis: computeFindingId({\n      analyst_id,\n      area: 'profile-quality',\n      subject,\n      claim: `richness:${richness.thin ? 'thin' : 'rich'}`,\n    }),\n  })\n}\n","/**\n *\n * `delegate` — the one generic delegation verb. You hand it an INTENT (what you want done) and it\n * hands that intent to a default AUTHORING supervisor: a router-brained supervisor whose standing\n * instruction is `supervisorInstructions()` (the authoring-agent-profiles skill). The supervisor\n * DECOMPOSES the intent and AUTHORS the worker profile it needs per sub-task — there is NO hardcoded\n * coder/researcher profile here. That is the whole point: `delegate('fix the failing test', …)` and\n * `delegate('research X and cite sources', …)` route through the SAME front door; the supervisor\n * writes a code-shaped or research-shaped worker on its own.\n *\n * It is a thin wrapper over `supervise()` — the one front door — so the conserved-budget pool, the\n * completion oracle (`deliverable`), the coordination toolbox, and equal-compute accounting all come\n * for free; nothing is hand-rolled. The result is `supervise()`'s `SupervisedResult` returned\n * UNCHANGED, so its `spentTotal` (`{ iterations, tokens, usd, ms }`) rides straight back to the\n * caller on BOTH paths — a `winner` carries the delivered worker's spend, a `no-winner` carries the\n * spend incurred before it failed. That cost channel means a `delegate()` caller always learns what\n * the delegation actually spent.\n *\n * @experimental\n */\n\nimport { ConfigError } from '../../errors'\nimport type { RouterTransportConfig } from '../router-http-transport'\nimport type { DeliverableSpec } from './completion-gate'\nimport type { ExecutorConfig } from './runtime'\nimport { supervise } from './supervise'\nimport type { SupervisorProfile } from './supervisor-agent'\nimport type { Budget, SupervisedResult } from './types'\n\n/** The conserved pool a `delegate()` call applies when the caller does not pass its own `budget`.\n *  A modest token ceiling + a small iteration ceiling — generous enough for a few-worker decompose,\n *  bounded enough that an unsupervised intent cannot run away. Callers override via `opts.budget`. */\nexport const defaultDelegateBudget: Budget = { maxIterations: 50, maxTokens: 200_000 }\n\n/** Inputs to {@link delegate}. The intent is the first positional arg; everything here is optional\n *  with explicit execution identity, so the common call names one exact supervisor profile. */\nexport interface DelegateOptions<Out = unknown> {\n  /** The completion oracle (settled ⟺ delivered) the authored workers settle against. Strongly\n   *  recommended — without it the supervisor trusts a worker's self-report. For a code intent,\n   *  `patchDelivered()` is the canonical example; for a free-form answer, a content check. */\n  readonly deliverable?: DeliverableSpec<Out>\n  /** WHERE the authored workers run — the worker-execution backend (`router-tools` / `sandbox` /\n   *  `cli-worktree` / …). The supervisor authors the worker PROFILE; this is the substrate it runs\n   *  on. Provide this OR `makeWorkerAgent`-style wiring through `supervise()` is unavailable. */\n  readonly backend?: ExecutorConfig\n  /** The conserved compute pool for the whole delegation. Defaults to {@link defaultDelegateBudget}. */\n  readonly budget?: Budget\n  /** Exact executable authoring supervisor. Model, prompt, harness, and provider live here. */\n  readonly supervisorProfile: SupervisorProfile\n  /** Router endpoint/auth for a `cli-base` supervisor; contains no behavioral settings. */\n  readonly router: RouterTransportConfig\n  /** Restrict the run to this subset of models (forwarded to `supervise()`). */\n  readonly allowedModels?: readonly string[]\n  readonly runId?: string\n}\n\n/**\n * Delegate an INTENT to a default authoring supervisor and return its `SupervisedResult` unchanged.\n *\n * The supervisor authors + spawns whatever worker the intent needs over the conserved-budget pool;\n * `result.spentTotal` reports what the whole delegation actually cost. A `winner` result carries the\n * authored worker's delivered output; a `no-winner` result names why (never a fabricated success).\n */\nexport async function delegate<Out = unknown>(\n  intent: string,\n  opts: DelegateOptions<Out>,\n): Promise<SupervisedResult<Out>> {\n  if (typeof intent !== 'string' || intent.trim().length === 0) {\n    throw new ConfigError('delegate: `intent` must be a non-empty string')\n  }\n  return supervise(opts.supervisorProfile, intent, {\n    budget: opts.budget ?? defaultDelegateBudget,\n    ...(opts.backend ? { backend: opts.backend } : {}),\n    ...(opts.deliverable ? { deliverable: opts.deliverable as DeliverableSpec<unknown> } : {}),\n    router: opts.router,\n    ...(opts.allowedModels ? { allowedModels: opts.allowedModels } : {}),\n    ...(opts.runId ? { runId: opts.runId } : {}),\n  }) as Promise<SupervisedResult<Out>>\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;AA+BA,SAAgB,kBAAkB,KAAsC;CACtE,MAAM,SAAS,mBAAmB,UAAU,GAAG;CAC/C,IAAI,CAAC,OAAO,SAAS,OAAO;CAC5B,MAAM,eAAe,OAAO,KAAK,QAAQ;CACzC,IAAI,OAAO,iBAAiB,YAAY,aAAa,KAAK,CAAC,CAAC,WAAW,GAAG,OAAO;CACjF,OAAO;EACL,GAAG,OAAO;EACV,MACE,OAAO,OAAO,KAAK,SAAS,YAAY,OAAO,KAAK,KAAK,SAAS,IAC9D,OAAO,KAAK,OACZ;EACN,QAAQ;GAAE,GAAG,OAAO,KAAK;GAAQ;EAAa;CAChD;AACF;;;AAIA,SAAgB,uBAAuB,MAAkC;CACvE,OAAO;EACL;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA,GAAI,MAAM,OAAO,CAAC,IAAI,aAAa,KAAK,MAAM,IAAI,CAAC;CACrD,CAAC,CAAC,KAAK,IAAI;AACb;;AAqBA,MAAa,mCAA8D;CACzE,sBAAsB;CACtB,sBAAsB;AACxB;;;AA0BA,SAAS,oBAAoB,SAA+B;CAC1D,MAAM,KAAM,QAAiC;CAC7C,IAAI,OAAO,OAAO,UAAU,OAAO;CACnC,IAAI,MAAM,OAAO,OAAO,UAAU;EAChC,MAAM,IAAI;EACV,IAAI,OAAO,EAAE,iBAAiB,UAAU,OAAO,EAAE;EACjD,IAAI,OAAO,EAAE,WAAW,UAAU,OAAO,EAAE;CAC7C;CACA,OAAO;AACT;;;AAIA,SAAgB,sBACd,SACA,MACiB;CACjB,MAAM,KAAK;EAAE,GAAG;EAAkC,GAAI,MAAM,cAAc,CAAC;CAAG;CAC9E,MAAM,eAAe,oBAAoB,OAAO;CAChD,MAAM,UAAU,aAAa,KAAK;CAClC,MAAM,oBAAoB,QAAQ;CAClC,MAAM,oBAAoB,UACtB,QAAQ,MAAM,IAAI,CAAC,CAAC,QAAQ,MAAM,EAAE,KAAK,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,SACvD;CACJ,MAAM,gBAAgB,WACjB,QAAQ,MAAM,cAAc,KAAK,CAAC,EAAA,CAAG,WAAW,UAAU,IAAI,KAC/D;CACJ,MAAM,iBACJ,OAAO,QAAQ,gBAAgB,YAAY,QAAQ,YAAY,KAAK,CAAC,CAAC,SAAS;CACjF,MAAM,QAAS,QAAgD;CAC/D,MAAM,WAAW,CAAC,CAAC,SAAS,OAAO,KAAK,KAAK,CAAC,CAAC,SAAS;CACxD,MAAM,SAAU,QAAQ,WAAkD;CAC1E,MAAM,YAAY,MAAM,QAAQ,MAAM,KAAK,OAAO,SAAS;CAC3D,MAAM,MAAO,QAA8C;CAC3D,MAAM,SAAS,CAAC,CAAC,OAAO,OAAO,KAAK,GAAG,CAAC,CAAC,SAAS;CAClD,MAAM,YAAa,QAAoD;CACvE,MAAM,eAAe,CAAC,CAAC,aAAa,OAAO,KAAK,SAAS,CAAC,CAAC,SAAS;CAEpE,MAAM,UAAoB,CAAC;CAC3B,MAAM,aACJ,oBAAoB,GAAG,wBAAwB,oBAAoB,GAAG;CACxE,IAAI,YACF,QAAQ,KACN,0BAA0B,kBAAkB,UAAU,kBAAkB,gBAAgB,GAAG,qBAAqB,cAAc,GAAG,qBAAqB,QACxJ;CACF,IAAI,CAAC,UACH,QAAQ,KAAK,yEAAyE;CACxF,IAAI,CAAC,WAAW,QAAQ,KAAK,wDAAwD;CACrF,IAAI,MAAM,YAAY,CAAC,QAAQ,QAAQ,KAAK,oDAAoD;CAGhG,MAAM,UAAU;EAAC,CAAC;EAAY;EAAU;EAAW;EAAgB,MAAM,WAAW,SAAS;CAAI;CACjG,MAAM,WAAW,QAAQ,OAAO,OAAO,CAAC,CAAC,SAAS,QAAQ;CAE1D,MAAM,OAAO,cAAe,CAAC,YAAY,CAAC,aAAa,CAAC;CAExD,OAAO;EACL,MAAM,QAAQ,QAAQ;EACtB;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;CACF;AACF;;;;AAKA,SAAgB,uBACd,UACA,MACgB;CAChB,MAAM,aAAa,MAAM,aAAa;CACtC,MAAM,UAAU,SAAS;CACzB,MAAM,QAAQ,SAAS,OACnB,WAAW,SAAS,KAAK,oCAAoC,SAAS,QAAQ,KAAK,IAAI,EAAE,KACzF,WAAW,SAAS,KAAK,8CAA8C,SAAS,WAAW,IAAA,CAAK,QAAQ,CAAC,EAAE;CAM/G,OAAO,YAAY;EACjB;EACA,UAP2C,SAAS,OAClD,SAAS,WAAW,MAClB,SACA,WACF;EAIF,MAAM;EACN;EACA;EACA,YAAY;EACZ,eAAe,CACb;GACE,MAAM;GACN,KAAK,WAAW;GAChB,SAAS,SAAS,SAAS,kBAAkB,SAAS,SAAS,kBAAkB,SAAS,SAAS,SAAS,UAAU,SAAS,UAAU,OAAO,SAAS,OAAO,YAAY,SAAS,SAAS,QAAQ,CAAC;EACzM,CACF;EACA,GAAI,SAAS,OACT,EAAE,oBAAoB,cAAc,SAAS,KAAK,UAAU,SAAS,QAAQ,KAAK,IAAI,EAAE,GAAG,IAC3F,CAAC;EACL,UAAU,iBAAiB;GACzB;GACA,MAAM;GACN;GACA,OAAO,YAAY,SAAS,OAAO,SAAS;EAC9C,CAAC;CACH,CAAC;AACH;;;;;;;;;;;;;;;;;;;;;;;;;;AC5MA,MAAa,wBAAgC;CAAE,eAAe;CAAI,WAAW;AAAQ;;;;;;;;AA+BrF,eAAsB,SACpB,QACA,MACgC;CAChC,IAAI,OAAO,WAAW,YAAY,OAAO,KAAK,CAAC,CAAC,WAAW,GACzD,MAAM,IAAI,YAAY,+CAA+C;CAEvE,OAAO,UAAU,KAAK,mBAAmB,QAAQ;EAC/C,QAAQ,KAAK,UAAU;EACvB,GAAI,KAAK,UAAU,EAAE,SAAS,KAAK,QAAQ,IAAI,CAAC;EAChD,GAAI,KAAK,cAAc,EAAE,aAAa,KAAK,YAAwC,IAAI,CAAC;EACxF,QAAQ,KAAK;EACb,GAAI,KAAK,gBAAgB,EAAE,eAAe,KAAK,cAAc,IAAI,CAAC;EAClE,GAAI,KAAK,QAAQ,EAAE,OAAO,KAAK,MAAM,IAAI,CAAC;CAC5C,CAAC;AACH"}
+{"version":3,"file":"delegate-Cp5-PsaX.js","names":[],"sources":["../src/runtime/supervise/authoring.ts","../src/runtime/supervise/delegate.ts"],"sourcesContent":["/**\n *\n * The supervisor's intelligence is AUTHORING the agents it spawns — not pressing buttons.\n *\n * Every agent here is three things: instructions (system prompt), tools, and a model — its\n * `AgentProfile`. The supervisor's job is to WRITE those profiles: read the task, decompose it,\n * and for each sub-task author a tailored worker recipe. `supervisorInstructions` is the how-to the\n * supervisor reads; canonical Runtime executors materialize the resulting profile.\n *\n * The skill is the single OPTIMIZABLE surface: edit it → the supervisor designs better agents.\n * That is the self-improvement lever (the prompt/skill lever), not the execution plumbing.\n *\n * @experimental\n */\n\nimport { type AnalystFinding, computeFindingId, makeFinding } from '@tangle-network/agent-eval'\nimport {\n  type AgentProfile,\n  type AgentProfilePrompt,\n  agentProfileSchema,\n} from '@tangle-network/agent-interface'\n\n/** What the supervisor AUTHORS per sub-task: one complete canonical profile whose name and\n *  task-specific system prompt are present. Every other `AgentProfile` axis is preserved exactly. */\nexport type AuthoredProfile = AgentProfile & {\n  readonly name: string\n  readonly prompt: AgentProfilePrompt & { readonly systemPrompt: string }\n}\n\n/** Narrow an untyped `spawn_worker` profile argument to an `AuthoredProfile`, or null if the\n *  supervisor failed to author one (empty/placeholder profile — a skill violation worth catching). */\nexport function asAuthoredProfile(raw: unknown): AuthoredProfile | null {\n  const parsed = agentProfileSchema.safeParse(raw)\n  if (!parsed.success) return null\n  const systemPrompt = parsed.data.prompt?.systemPrompt\n  if (typeof systemPrompt !== 'string' || systemPrompt.trim().length === 0) return null\n  return {\n    ...parsed.data,\n    name:\n      typeof parsed.data.name === 'string' && parsed.data.name.length > 0\n        ? parsed.data.name\n        : 'worker',\n    prompt: { ...parsed.data.prompt, systemPrompt },\n  }\n}\n\n/** The supervisor skill: an explicit profile-authoring instruction, never an implicit Runtime\n * policy. Editing this text changes how a profile designs the descendants it spawns. */\nexport function supervisorInstructions(opts?: { goal?: string }): string {\n  return [\n    'Your delegation craft is AUTHORING: a spawned worker is exactly as good as the profile you write.',\n    '',\n    'For the task you are given:',\n    '1. DECOMPOSE it into the smallest set of sub-tasks a single focused worker can each deliver.',\n    '2. For EACH sub-task, AUTHOR a worker by calling spawn_worker with a COMPLETE `profile`:',\n    '   • name and description: who this specialist is and why it exists.',\n    '   • prompt.systemPrompt: rich instructions for THIS sub-task — exact output, process, evidence, and what \"done\" means.',\n    '   • model.default, model.reasoningEffort, and harness: choose the execution system deliberately when the task benefits from it.',\n    '   • tools, mcp, resources.skills/files/instructions, hooks, subagents, permissions, and modes: grant or attach every capability the worker needs; omit an axis only when it is intentionally unnecessary.',\n    '   • tools.agent_runtime_coordination_spawn_worker: true ONLY when this child should author and drive descendants. Add only the other agent_runtime_coordination_<verb> tools it will call, such as observe_agent or steer_agent.',\n    '   • A child with spawn_worker MUST carry this complete profile-authoring instruction as an immutable resources.skills entry with resources.failOnError: true, so it can author its own descendants from the same contract.',\n    '   • metadata may describe the work, but it never grants recursion or selects a Runtime execution path.',\n    '   NEVER spawn a worker with an empty profile. The quality of the worker IS the quality of the profile you write.',\n    '3. End your turn while your workers run: Runtime wakes you with each settled worker. A settlement says valid:true only if the deployable check passed.',\n    '4. If a worker did NOT deliver, AUTHOR A NEW profile whose prompt.systemPrompt names the SPECIFIC failure and how to fix it — never just retry the same profile.',\n    \"5. read_journal to re-read YOUR OWN record before you decide the next move: every spawn you made, every settle, every question and answer, every steer, every analyst finding — oldest first, this node only, including what you did before a restart. Use it to see what you already tried instead of trying it again. It is paged: pass the returned nextRow as the next call's sinceRow, narrow with kinds, and raise limit/maxBytes only as far as you will actually read. A truncated:true page means a bound cut it short — keep paging before you conclude you have read everything.\",\n    '6. AUTHOR YOUR OWN LENS when the questions you can already ask of a settled trace do not cover the failure you are chasing: define_analyst takes an id, a description, an area, the question in your own words, the instructions for answering it with trace evidence, and the smallest toolGroup that can answer it (model is the seat it runs on; omit it for the run default). It is DATA, never code. Then run_analyst it on any settled worker like a lens the run shipped with, and read the finding. list_analysts shows what you have. Define a lens when you need a different question asked — not a second copy of a question already on the menu.',\n    '7. EVERY refusal you get back carries a `reason` naming the exact unmet condition. Read it and change that condition — a spawn refused for budget-exhausted needs a smaller budget or the refund of a settled worker, an invalid-profile needs the named field fixed, a submit_result refused because the check THREW is a broken check to report, not a result to resubmit. Never repeat a call that was refused without changing what it was refused for.',\n    '8. ask_parent ONLY when you genuinely cannot decide, and then READ ITS OUTCOME. \"queued-for-parent\" means an inbox above you now holds the question. \"no-parent\" means no inbox above you is configured to receive it: the question is still on the run record for anyone watching, but nothing will route an answer back to you, so do not block. Decide it with answer_question, or answer_question with deferReason to record that it stays open, and carry on — a blocking question left undecided also refuses your stop.',\n    '9. Once the work is delivered, end your turn with nothing left running.',\n    ...(opts?.goal ? ['', `The goal: ${opts.goal}`] : []),\n  ].join('\\n')\n}\n\n// ── Profile-richness gate ────────────────────────────────────────────────────\n//\n// The supervisor's product is the worker PROFILE it authors. The failure mode the existing\n// gates miss: `asAuthoredProfile` / `local-harness` only reject a FULLY EMPTY system prompt —\n// a two-sentence stub passes. `assessAuthoredProfile` OBSERVES the authored artifact (it reads\n// no judge verdict, so it steers cleanly past `assertTraceDerivedFindings`) and flags THIN:\n// a short/few-line system prompt, OR no tools, OR no skills, OR no MCP when the task needs one.\n// It emits a real `AnalystFinding` so it rides the SAME coordination bus the driver pulls via\n// receives in its next wake — the supervisor can self-correct and re-author richer.\n\n/** Thresholds below which a system prompt is treated as a thin stub. Tunable per call. */\nexport interface ProfileRichnessThresholds {\n  /** A prompt shorter than this many characters is thin (default 600). */\n  readonly minSystemPromptChars: number\n  /** A prompt with fewer than this many non-blank lines is thin (default 6). */\n  readonly minSystemPromptLines: number\n}\n\n/** Default thresholds for `ProfileRichnessThresholds` — 600 chars / 6 lines minimum system prompt. */\nexport const defaultProfileRichnessThresholds: ProfileRichnessThresholds = {\n  minSystemPromptChars: 600,\n  minSystemPromptLines: 6,\n}\n\n/** Per-field verdict on one authored profile — the raw material the bench renders + scores. */\nexport interface ProfileRichness {\n  readonly name: string\n  /** The resolved system prompt (canonical `prompt.systemPrompt`, the sandbox `prompt.system`\n   *  convention, or a bare-string prompt — whichever the author used). */\n  readonly systemPrompt: string\n  readonly systemPromptChars: number\n  readonly systemPromptLines: number\n  readonly sentenceCount: number\n  readonly hasDescription: boolean\n  readonly hasTools: boolean\n  readonly hasSkills: boolean\n  readonly hasMcp: boolean\n  readonly hasSubagents: boolean\n  /** 0..1 — fraction of richness signals present (prompt-depth + the four levers). */\n  readonly richness: number\n  /** True when the supervisor authored a stub instead of a real profile. */\n  readonly thin: boolean\n  /** The specific reasons it is thin (empty when rich) — used in the finding's action. */\n  readonly reasons: string[]\n}\n\n/** Read the system prompt from any authored shape: canonical `prompt.systemPrompt`, the sandbox\n *  `prompt.system` convention, or a bare-string `prompt`. */\nfunction resolveSystemPrompt(profile: AgentProfile): string {\n  const pr = (profile as { prompt?: unknown }).prompt\n  if (typeof pr === 'string') return pr\n  if (pr && typeof pr === 'object') {\n    const o = pr as { systemPrompt?: unknown; system?: unknown }\n    if (typeof o.systemPrompt === 'string') return o.systemPrompt\n    if (typeof o.system === 'string') return o.system\n  }\n  return ''\n}\n\n/** OBSERVE one authored `AgentProfile` and score its richness (no judge verdict is read). The task\n *  context (`needsMcp`) lets a domain say \"this work needs a data/tool MCP\" so a missing MCP counts. */\nexport function assessAuthoredProfile(\n  profile: AgentProfile,\n  opts?: { needsMcp?: boolean; thresholds?: Partial<ProfileRichnessThresholds> },\n): ProfileRichness {\n  const th = { ...defaultProfileRichnessThresholds, ...(opts?.thresholds ?? {}) }\n  const systemPrompt = resolveSystemPrompt(profile)\n  const trimmed = systemPrompt.trim()\n  const systemPromptChars = trimmed.length\n  const systemPromptLines = trimmed\n    ? trimmed.split('\\n').filter((l) => l.trim().length > 0).length\n    : 0\n  const sentenceCount = trimmed\n    ? (trimmed.match(/[.!?](\\s|$)/g) ?? []).length || (trimmed ? 1 : 0)\n    : 0\n  const hasDescription =\n    typeof profile.description === 'string' && profile.description.trim().length > 0\n  const tools = (profile as { tools?: Record<string, unknown> }).tools\n  const hasTools = !!tools && Object.keys(tools).length > 0\n  const skills = (profile.resources as { skills?: unknown[] } | undefined)?.skills\n  const hasSkills = Array.isArray(skills) && skills.length > 0\n  const mcp = (profile as { mcp?: Record<string, unknown> }).mcp\n  const hasMcp = !!mcp && Object.keys(mcp).length > 0\n  const subagents = (profile as { subagents?: Record<string, unknown> }).subagents\n  const hasSubagents = !!subagents && Object.keys(subagents).length > 0\n\n  const reasons: string[] = []\n  const promptThin =\n    systemPromptChars < th.minSystemPromptChars || systemPromptLines < th.minSystemPromptLines\n  if (promptThin)\n    reasons.push(\n      `system prompt is thin (${systemPromptChars} chars, ${systemPromptLines} lines; need ≥${th.minSystemPromptChars} chars and ≥${th.minSystemPromptLines} lines)`,\n    )\n  if (!hasTools)\n    reasons.push('no tools granted (a worker can only act through the tools you grant it)')\n  if (!hasSkills) reasons.push('no skills attached (no reusable how-to notes injected)')\n  if (opts?.needsMcp && !hasMcp) reasons.push('no MCP server, but the task needs data/tool access')\n\n  // Richness = fraction of signals present. Prompt-depth is one signal; the four levers are the rest.\n  const signals = [!promptThin, hasTools, hasSkills, hasDescription, opts?.needsMcp ? hasMcp : true]\n  const richness = signals.filter(Boolean).length / signals.length\n  // THIN ⟺ the prompt is a stub OR the worker has no levers at all (no tools AND no skills AND no mcp).\n  const thin = promptThin || (!hasTools && !hasSkills && !hasMcp)\n\n  return {\n    name: profile.name ?? 'worker',\n    systemPrompt,\n    systemPromptChars,\n    systemPromptLines,\n    sentenceCount,\n    hasDescription,\n    hasTools,\n    hasSkills,\n    hasMcp,\n    hasSubagents,\n    richness,\n    thin,\n    reasons,\n  }\n}\n\n/** Turn a {@link ProfileRichness} verdict into a bus-routable `AnalystFinding` (area `profile-quality`).\n *  Severity scales with thinness; the recommended action names the MISSING lever so the supervisor can\n *  re-author. `subject` = the worker name so per-worker findings diff cleanly across re-authors. */\nexport function profileRichnessFinding(\n  richness: ProfileRichness,\n  opts?: { analystId?: string; runId?: string },\n): AnalystFinding {\n  const analyst_id = opts?.analystId ?? 'profile-richness'\n  const subject = richness.name\n  const claim = richness.thin\n    ? `Worker \"${richness.name}\" was authored as a THIN profile: ${richness.reasons.join('; ')}.`\n    : `Worker \"${richness.name}\" was authored as a rich profile (richness ${(richness.richness * 100).toFixed(0)}%).`\n  const severity: AnalystFinding['severity'] = richness.thin\n    ? richness.richness < 0.25\n      ? 'high'\n      : 'medium'\n    : 'info'\n  return makeFinding({\n    analyst_id,\n    severity,\n    area: 'profile-quality',\n    claim,\n    subject,\n    confidence: 0.9,\n    evidence_refs: [\n      {\n        kind: 'metric',\n        uri: `profile:${subject}`,\n        excerpt: `chars=${richness.systemPromptChars} lines=${richness.systemPromptLines} tools=${richness.hasTools} skills=${richness.hasSkills} mcp=${richness.hasMcp} richness=${richness.richness.toFixed(2)}`,\n      },\n    ],\n    ...(richness.thin\n      ? { recommended_action: `Re-author \"${richness.name}\" with: ${richness.reasons.join('; ')}.` }\n      : {}),\n    id_basis: computeFindingId({\n      analyst_id,\n      area: 'profile-quality',\n      subject,\n      claim: `richness:${richness.thin ? 'thin' : 'rich'}`,\n    }),\n  })\n}\n","/**\n *\n * `delegate` — the one generic delegation verb. You hand it an INTENT (what you want done) and it\n * hands that intent to a default AUTHORING supervisor: a router-brained supervisor whose standing\n * instruction is `supervisorInstructions()` (the authoring-agent-profiles skill). The supervisor\n * DECOMPOSES the intent and AUTHORS the worker profile it needs per sub-task — there is NO hardcoded\n * coder/researcher profile here. That is the whole point: `delegate('fix the failing test', …)` and\n * `delegate('research X and cite sources', …)` route through the SAME front door; the supervisor\n * writes a code-shaped or research-shaped worker on its own.\n *\n * It is a thin wrapper over `supervise()` — the one front door — so the conserved-budget pool, the\n * completion oracle (`deliverable`), the coordination toolbox, and equal-compute accounting all come\n * for free; nothing is hand-rolled. The result is `supervise()`'s `SupervisedResult` returned\n * UNCHANGED, so its `spentTotal` (`{ iterations, tokens, usd, ms }`) rides straight back to the\n * caller on BOTH paths — a `winner` carries the delivered worker's spend, a `no-winner` carries the\n * spend incurred before it failed. That cost channel means a `delegate()` caller always learns what\n * the delegation actually spent.\n *\n * @experimental\n */\n\nimport { ConfigError } from '../../errors'\nimport type { RouterTransportConfig } from '../router-http-transport'\nimport type { DeliverableSpec } from './completion-gate'\nimport type { ExecutorConfig } from './runtime'\nimport { supervise } from './supervise'\nimport type { SupervisorProfile } from './supervisor-agent'\nimport type { Budget, SupervisedResult } from './types'\n\n/** The conserved pool a `delegate()` call applies when the caller does not pass its own `budget`.\n *  A modest token ceiling + a small iteration ceiling — generous enough for a few-worker decompose,\n *  bounded enough that an unsupervised intent cannot run away. Callers override via `opts.budget`. */\nexport const defaultDelegateBudget: Budget = { maxIterations: 50, maxTokens: 200_000 }\n\n/** Inputs to {@link delegate}. The intent is the first positional arg; everything here is optional\n *  with explicit execution identity, so the common call names one exact supervisor profile. */\nexport interface DelegateOptions<Out = unknown> {\n  /** The completion oracle (settled ⟺ delivered) the authored workers settle against. Strongly\n   *  recommended — without it the supervisor trusts a worker's self-report. For a code intent,\n   *  `patchDelivered()` is the canonical example; for a free-form answer, a content check. */\n  readonly deliverable?: DeliverableSpec<Out>\n  /** WHERE the authored workers run — the worker-execution backend (`router-tools` / `sandbox` /\n   *  `cli-worktree` / …). The supervisor authors the worker PROFILE; this is the substrate it runs\n   *  on. Provide this OR `makeWorkerAgent`-style wiring through `supervise()` is unavailable. */\n  readonly backend?: ExecutorConfig\n  /** The conserved compute pool for the whole delegation. Defaults to {@link defaultDelegateBudget}. */\n  readonly budget?: Budget\n  /** Exact executable authoring supervisor. Model, prompt, harness, and provider live here. */\n  readonly supervisorProfile: SupervisorProfile\n  /** Router endpoint/auth for a `cli-base` supervisor; contains no behavioral settings. */\n  readonly router: RouterTransportConfig\n  /** Restrict the run to this subset of models (forwarded to `supervise()`). */\n  readonly allowedModels?: readonly string[]\n  readonly runId?: string\n}\n\n/**\n * Delegate an INTENT to a default authoring supervisor and return its `SupervisedResult` unchanged.\n *\n * The supervisor authors + spawns whatever worker the intent needs over the conserved-budget pool;\n * `result.spentTotal` reports what the whole delegation actually cost. A `winner` result carries the\n * authored worker's delivered output; a `no-winner` result names why (never a fabricated success).\n */\nexport async function delegate<Out = unknown>(\n  intent: string,\n  opts: DelegateOptions<Out>,\n): Promise<SupervisedResult<Out>> {\n  if (typeof intent !== 'string' || intent.trim().length === 0) {\n    throw new ConfigError('delegate: `intent` must be a non-empty string')\n  }\n  return supervise(opts.supervisorProfile, intent, {\n    budget: opts.budget ?? defaultDelegateBudget,\n    ...(opts.backend ? { backend: opts.backend } : {}),\n    ...(opts.deliverable ? { deliverable: opts.deliverable as DeliverableSpec<unknown> } : {}),\n    router: opts.router,\n    ...(opts.allowedModels ? { allowedModels: opts.allowedModels } : {}),\n    ...(opts.runId ? { runId: opts.runId } : {}),\n  }) as Promise<SupervisedResult<Out>>\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;AA+BA,SAAgB,kBAAkB,KAAsC;CACtE,MAAM,SAAS,mBAAmB,UAAU,GAAG;CAC/C,IAAI,CAAC,OAAO,SAAS,OAAO;CAC5B,MAAM,eAAe,OAAO,KAAK,QAAQ;CACzC,IAAI,OAAO,iBAAiB,YAAY,aAAa,KAAK,CAAC,CAAC,WAAW,GAAG,OAAO;CACjF,OAAO;EACL,GAAG,OAAO;EACV,MACE,OAAO,OAAO,KAAK,SAAS,YAAY,OAAO,KAAK,KAAK,SAAS,IAC9D,OAAO,KAAK,OACZ;EACN,QAAQ;GAAE,GAAG,OAAO,KAAK;GAAQ;EAAa;CAChD;AACF;;;AAIA,SAAgB,uBAAuB,MAAkC;CACvE,OAAO;EACL;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA,GAAI,MAAM,OAAO,CAAC,IAAI,aAAa,KAAK,MAAM,IAAI,CAAC;CACrD,CAAC,CAAC,KAAK,IAAI;AACb;;AAqBA,MAAa,mCAA8D;CACzE,sBAAsB;CACtB,sBAAsB;AACxB;;;AA0BA,SAAS,oBAAoB,SAA+B;CAC1D,MAAM,KAAM,QAAiC;CAC7C,IAAI,OAAO,OAAO,UAAU,OAAO;CACnC,IAAI,MAAM,OAAO,OAAO,UAAU;EAChC,MAAM,IAAI;EACV,IAAI,OAAO,EAAE,iBAAiB,UAAU,OAAO,EAAE;EACjD,IAAI,OAAO,EAAE,WAAW,UAAU,OAAO,EAAE;CAC7C;CACA,OAAO;AACT;;;AAIA,SAAgB,sBACd,SACA,MACiB;CACjB,MAAM,KAAK;EAAE,GAAG;EAAkC,GAAI,MAAM,cAAc,CAAC;CAAG;CAC9E,MAAM,eAAe,oBAAoB,OAAO;CAChD,MAAM,UAAU,aAAa,KAAK;CAClC,MAAM,oBAAoB,QAAQ;CAClC,MAAM,oBAAoB,UACtB,QAAQ,MAAM,IAAI,CAAC,CAAC,QAAQ,MAAM,EAAE,KAAK,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,SACvD;CACJ,MAAM,gBAAgB,WACjB,QAAQ,MAAM,cAAc,KAAK,CAAC,EAAA,CAAG,WAAW,UAAU,IAAI,KAC/D;CACJ,MAAM,iBACJ,OAAO,QAAQ,gBAAgB,YAAY,QAAQ,YAAY,KAAK,CAAC,CAAC,SAAS;CACjF,MAAM,QAAS,QAAgD;CAC/D,MAAM,WAAW,CAAC,CAAC,SAAS,OAAO,KAAK,KAAK,CAAC,CAAC,SAAS;CACxD,MAAM,SAAU,QAAQ,WAAkD;CAC1E,MAAM,YAAY,MAAM,QAAQ,MAAM,KAAK,OAAO,SAAS;CAC3D,MAAM,MAAO,QAA8C;CAC3D,MAAM,SAAS,CAAC,CAAC,OAAO,OAAO,KAAK,GAAG,CAAC,CAAC,SAAS;CAClD,MAAM,YAAa,QAAoD;CACvE,MAAM,eAAe,CAAC,CAAC,aAAa,OAAO,KAAK,SAAS,CAAC,CAAC,SAAS;CAEpE,MAAM,UAAoB,CAAC;CAC3B,MAAM,aACJ,oBAAoB,GAAG,wBAAwB,oBAAoB,GAAG;CACxE,IAAI,YACF,QAAQ,KACN,0BAA0B,kBAAkB,UAAU,kBAAkB,gBAAgB,GAAG,qBAAqB,cAAc,GAAG,qBAAqB,QACxJ;CACF,IAAI,CAAC,UACH,QAAQ,KAAK,yEAAyE;CACxF,IAAI,CAAC,WAAW,QAAQ,KAAK,wDAAwD;CACrF,IAAI,MAAM,YAAY,CAAC,QAAQ,QAAQ,KAAK,oDAAoD;CAGhG,MAAM,UAAU;EAAC,CAAC;EAAY;EAAU;EAAW;EAAgB,MAAM,WAAW,SAAS;CAAI;CACjG,MAAM,WAAW,QAAQ,OAAO,OAAO,CAAC,CAAC,SAAS,QAAQ;CAE1D,MAAM,OAAO,cAAe,CAAC,YAAY,CAAC,aAAa,CAAC;CAExD,OAAO;EACL,MAAM,QAAQ,QAAQ;EACtB;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;CACF;AACF;;;;AAKA,SAAgB,uBACd,UACA,MACgB;CAChB,MAAM,aAAa,MAAM,aAAa;CACtC,MAAM,UAAU,SAAS;CACzB,MAAM,QAAQ,SAAS,OACnB,WAAW,SAAS,KAAK,oCAAoC,SAAS,QAAQ,KAAK,IAAI,EAAE,KACzF,WAAW,SAAS,KAAK,8CAA8C,SAAS,WAAW,IAAA,CAAK,QAAQ,CAAC,EAAE;CAM/G,OAAO,YAAY;EACjB;EACA,UAP2C,SAAS,OAClD,SAAS,WAAW,MAClB,SACA,WACF;EAIF,MAAM;EACN;EACA;EACA,YAAY;EACZ,eAAe,CACb;GACE,MAAM;GACN,KAAK,WAAW;GAChB,SAAS,SAAS,SAAS,kBAAkB,SAAS,SAAS,kBAAkB,SAAS,SAAS,SAAS,UAAU,SAAS,UAAU,OAAO,SAAS,OAAO,YAAY,SAAS,SAAS,QAAQ,CAAC;EACzM,CACF;EACA,GAAI,SAAS,OACT,EAAE,oBAAoB,cAAc,SAAS,KAAK,UAAU,SAAS,QAAQ,KAAK,IAAI,EAAE,GAAG,IAC3F,CAAC;EACL,UAAU,iBAAiB;GACzB;GACA,MAAM;GACN;GACA,OAAO,YAAY,SAAS,OAAO,SAAS;EAC9C,CAAC;CACH,CAAC;AACH;;;;;;;;;;;;;;;;;;;;;;;;;;AC5MA,MAAa,wBAAgC;CAAE,eAAe;CAAI,WAAW;AAAQ;;;;;;;;AA+BrF,eAAsB,SACpB,QACA,MACgC;CAChC,IAAI,OAAO,WAAW,YAAY,OAAO,KAAK,CAAC,CAAC,WAAW,GACzD,MAAM,IAAI,YAAY,+CAA+C;CAEvE,OAAO,UAAU,KAAK,mBAAmB,QAAQ;EAC/C,QAAQ,KAAK,UAAU;EACvB,GAAI,KAAK,UAAU,EAAE,SAAS,KAAK,QAAQ,IAAI,CAAC;EAChD,GAAI,KAAK,cAAc,EAAE,aAAa,KAAK,YAAwC,IAAI,CAAC;EACxF,QAAQ,KAAK;EACb,GAAI,KAAK,gBAAgB,EAAE,eAAe,KAAK,cAAc,IAAI,CAAC;EAClE,GAAI,KAAK,QAAQ,EAAE,OAAO,KAAK,MAAM,IAAI,CAAC;CAC5C,CAAC;AACH"}

package/dist/durable.d.ts

@@ -1,4 +1,4 @@
-import { $l as PursuitVersions, Cu as FileObserverJournal, D as RetainedPendingCause, Dd as SelectSeat, Du as observerRecordDigest, Eu as ObserverRecordKind, Gf as CoordinationAuthentication, Hd as DeadlineCheckIn, Jf as CoordinationReadAccess, Jl as JudgedPursuitVersion, Km as ToolLoopCompactionOptions, Mv as DownMessageAuthorizationInput, Qf as CoordinationRejectReason, Ql as PursuitVersionStop, Rv as MapOnSettleRoute, Su as projectPursuit, Td as SeatSelectionInput, Tu as ObserverRecord, V as NodeId, Xl as NextPursuitVersionInput, Yf as CoordinationTransportOptions, Yl as NextPursuitVersion, Zf as CoordinationHttpAudit, Zl as PursuitVersionChain, _u as PursuitNodeUsage, au as SupervisePursuitError, bu as PursuitRunTotals, cd as AuthorizedSpawnContext, cu as superviseGraphPursuit, dd as WorkerSlots, df as ResolveDriveHarness, du as PursuitObserverDelivery, eu as SettledPursuitVersion, fu as PursuitCostProvenance, gu as PursuitNodeTiming, gv as EscalateQuestion, hu as PursuitNodeProjection, it as SupervisedResult, iu as SuperviseGraphPursuitOptions, jv as ContinuityMode, kv as AuthorizedDownMessage, lf as DriveHarnessOwnerContext, lu as supervisePursuit, mu as PursuitNodePlatform, nd as RUN_FORK_CORRELATION_KEYS, nf as readRootStream, nu as VersionVerdict, ou as SupervisePursuitOptions, ov as WorkerWatchOptions, pu as PursuitNodeCost, qf as CoordinationPublicAddress, rf as readRootStreamReceipt, rt as RootStreamReceipt, ru as assertPursuitVersions, sd as AuthorizedSpawn, su as SupervisedPursuitResult, td as PursuitFork, tf as RootStreamRecord, tu as VersionJudge, uu as RunReopen, vu as PursuitProjection, wd as SeatSelection, wu as ObserverJournal, wy as AnalyzeOnSettleRoute, xu as PursuitStatus, yf as ManagerWakePolicy, yu as PursuitRunProjection, yv as QuestionEscalationOutcome } from "./types-D0ZRqIJh.js";
+import { $f as CoordinationAuthentication, $l as NextPursuitVersionInput, Ad as SelectSeat, Au as observerRecordDigest, Bv as DownMessageAuthorizationInput, Cu as PursuitRunTotals, D as RetainedPendingCause, Dd as SeatSelection, Df as ManagerWakePolicy, Du as ObserverJournal, Dv as QuestionEscalationOutcome, Eu as FileObserverJournal, Gd as DeadlineCheckIn, Jd as RunParkingPolicy, Kd as ParkedRun, Kv as MapOnSettleRoute, Lv as AuthorizedDownMessage, My as AnalyzeOnSettleRoute, Od as SeatSelectionInput, Ou as ObserverRecord, Ql as NextPursuitVersion, Su as PursuitRunProjection, Tu as projectPursuit, V as NodeId, Xd as readRunParking, Yd as RunParkingRecord, Zl as JudgedPursuitVersion, _f as DriveHarnessOwnerContext, _u as PursuitNodePlatform, ad as RUN_FORK_CORRELATION_KEYS, ap as CoordinationHttpAudit, au as VersionVerdict, bu as PursuitNodeUsage, cu as SupervisePursuitError, dd as AuthorizedSpawnContext, df as readRootStreamReceipt, du as superviseGraphPursuit, eh as ToolLoopCompactionOptions, eu as PursuitVersionChain, fu as supervisePursuit, gu as PursuitNodeCost, hu as PursuitCostProvenance, id as PursuitFork, it as SupervisedResult, iu as VersionJudge, ku as ObserverRecordKind, lf as RootStreamRecord, lu as SupervisePursuitOptions, md as WorkerSlots, mu as PursuitObserverDelivery, mv as WorkerWatchOptions, np as CoordinationReadAccess, nu as PursuitVersions, op as CoordinationRejectReason, ou as assertPursuitVersions, pu as RunReopen, qd as RunParkedError, rp as CoordinationTransportOptions, rt as RootStreamReceipt, ru as SettledPursuitVersion, su as SuperviseGraphPursuitOptions, tp as CoordinationPublicAddress, tu as PursuitVersionStop, ud as AuthorizedSpawn, uf as readRootStream, uu as SupervisedPursuitResult, vu as PursuitNodeProjection, wu as PursuitStatus, wv as EscalateQuestion, xu as PursuitProjection, yf as ResolveDriveHarness, yu as PursuitNodeTiming, zv as ContinuityMode } from "./types-hc4cmSc7.js";
 import { n as RuntimeDecisionKind, r as RuntimeDecisionPoint, t as RuntimeDecisionEvidenceRef } from "./runtime-hooks-nsqxSUFL.js";
 import { LockIdentity } from "@tangle-network/agent-eval/lock-holder";
 //#region src/durable/chat-engine.d.ts
@@ -137,6 +137,8 @@ declare function deriveExecutionId(input: {
 /** What the lock file records about its holder. */
 interface RunDirectoryLockHolder {
   readonly pid: number;
+  /** Monotonic acquisition generation; absent on legacy lock records. */
+  readonly generation?: number;
   /** ISO instant the holder took the lock. */
   readonly startedAt: string;
   readonly runId: string;
@@ -151,7 +153,10 @@ interface RunDirectoryLockHolder {
 }
 /** A held lock. `release()` removes the file; it is safe to call more than once. */
 interface RunDirectoryLock extends RunDirectoryLockHolder {
+  readonly generation: number;
   readonly path: string;
+  /** Bind durable writes to this acquisition, including work queued before release. */
+  run<T>(operation: () => T): T;
   release(): Promise<void>;
 }
 /** The directory is held by a live process. `holder` is what that process recorded. */
@@ -261,5 +266,5 @@ interface DurableSupervisionDiscovery {
  */
 declare function discoverDurableSupervisionRun(runDir: string): Promise<DurableSupervisionDiscovery>;
 //#endregion
-export { type AnalyzeOnSettleRoute, type AuthorizedDownMessage, type AuthorizedSpawn, type AuthorizedSpawnContext, type ChatStreamEvent, type ChatTurnHooks, type ChatTurnIdentity, type ChatTurnProducer, type ChatTurnResult, type ContinuityMode, type CoordinationAuthentication, type CoordinationHttpAudit, type CoordinationPublicAddress, type CoordinationReadAccess, type CoordinationRejectReason, type CoordinationTransportOptions, type DeadlineCheckIn, type DownMessageAuthorizationInput, type DriveHarnessOwnerContext, type DurableCoordinationStreamIdentity, type DurableFailureRecord, type DurableSupervisionDiscovery, type EscalateQuestion, FileObserverJournal, type JudgedPursuitVersion, type ManagerWakePolicy, type MapOnSettleRoute, type NextPursuitVersion, type NextPursuitVersionInput, type ObserverJournal, type ObserverRecord, type ObserverRecordKind, type PursuitCostProvenance, type PursuitFork, type PursuitNodeCost, type PursuitNodePlatform, type PursuitNodeProjection, type PursuitNodeTiming, type PursuitNodeUsage, type PursuitObserverDelivery, type PursuitProjection, type PursuitRunProjection, type PursuitRunTotals, type PursuitStatus, type PursuitVersionChain, type PursuitVersionStop, type PursuitVersions, type QuestionEscalationOutcome, RUN_FORK_CORRELATION_KEYS, type ResolveDriveHarness, type RetainedPendingCause, type RootStreamRecord, type RunChatTurnInput, type RunDirectoryHolderLiveness, type RunDirectoryLock, type RunDirectoryLockHolder, RunDirectoryLockedError, type RunReopen, type RuntimeDecisionEvidenceRef, type RuntimeDecisionKind, type RuntimeDecisionPoint, type SeatSelection, type SeatSelectionInput, type SelectSeat, type SettledPursuitVersion, SettledRunDirectoryError, type SuperviseGraphPursuitOptions, SupervisePursuitError, type SupervisePursuitOptions, type SupervisedPursuitResult, type ToolLoopCompactionOptions, type VersionJudge, type VersionVerdict, type WorkerSlots, type WorkerWatchOptions, acquireRunDirectoryLock, assertPursuitVersions, deriveExecutionId, discoverDurableSupervisionRun, handleChatTurn, observerRecordDigest, projectPursuit, readFailureRecord, readRootStream, readRootStreamReceipt, readRunDirectoryLock, readSettleRecord, runDirectoryHolderIsLive, superviseGraphPursuit, supervisePursuit };
+export { type AnalyzeOnSettleRoute, type AuthorizedDownMessage, type AuthorizedSpawn, type AuthorizedSpawnContext, type ChatStreamEvent, type ChatTurnHooks, type ChatTurnIdentity, type ChatTurnProducer, type ChatTurnResult, type ContinuityMode, type CoordinationAuthentication, type CoordinationHttpAudit, type CoordinationPublicAddress, type CoordinationReadAccess, type CoordinationRejectReason, type CoordinationTransportOptions, type DeadlineCheckIn, type DownMessageAuthorizationInput, type DriveHarnessOwnerContext, type DurableCoordinationStreamIdentity, type DurableFailureRecord, type DurableSupervisionDiscovery, type EscalateQuestion, FileObserverJournal, type JudgedPursuitVersion, type ManagerWakePolicy, type MapOnSettleRoute, type NextPursuitVersion, type NextPursuitVersionInput, type ObserverJournal, type ObserverRecord, type ObserverRecordKind, type ParkedRun, type PursuitCostProvenance, type PursuitFork, type PursuitNodeCost, type PursuitNodePlatform, type PursuitNodeProjection, type PursuitNodeTiming, type PursuitNodeUsage, type PursuitObserverDelivery, type PursuitProjection, type PursuitRunProjection, type PursuitRunTotals, type PursuitStatus, type PursuitVersionChain, type PursuitVersionStop, type PursuitVersions, type QuestionEscalationOutcome, RUN_FORK_CORRELATION_KEYS, type ResolveDriveHarness, type RetainedPendingCause, type RootStreamRecord, type RunChatTurnInput, type RunDirectoryHolderLiveness, type RunDirectoryLock, type RunDirectoryLockHolder, RunDirectoryLockedError, RunParkedError, type RunParkingPolicy, type RunParkingRecord, type RunReopen, type RuntimeDecisionEvidenceRef, type RuntimeDecisionKind, type RuntimeDecisionPoint, type SeatSelection, type SeatSelectionInput, type SelectSeat, type SettledPursuitVersion, SettledRunDirectoryError, type SuperviseGraphPursuitOptions, SupervisePursuitError, type SupervisePursuitOptions, type SupervisedPursuitResult, type ToolLoopCompactionOptions, type VersionJudge, type VersionVerdict, type WorkerSlots, type WorkerWatchOptions, acquireRunDirectoryLock, assertPursuitVersions, deriveExecutionId, discoverDurableSupervisionRun, handleChatTurn, observerRecordDigest, projectPursuit, readFailureRecord, readRootStream, readRootStreamReceipt, readRunDirectoryLock, readRunParking, readSettleRecord, runDirectoryHolderIsLive, superviseGraphPursuit, supervisePursuit };
 //# sourceMappingURL=durable.d.ts.map

package/dist/durable.js

@@ -1,18 +1,18 @@
-import { a as syncDurableDirectory, i as publishExclusiveDurableFile, o as writeAtomicDurableFile } from "./durable-file-DLcrzrjL.js";
-import { Bi as assertValidBudget, C as FileSpawnJournal, Mi as contentAddress, Qr as authoredProfileDigest, _t as readCommittedJsonLines, ct as loadSpawnForest, hr as isFiniteNumber, ji as detachedSnapshot, mr as failedItems, mt as isNoEntError, pt as appendCommittedJsonLine, r as assertResumeLimits } from "./supervisor-D3M9ab1v.js";
+import { a as publishExclusiveDurableFile, c as captureRunWriteContext, d as runGenerationPath, f as runWithWriteClaim, l as fenceRunWrite, o as syncDurableDirectory, s as writeAtomicDurableFile, t as spawnJournalRecordEvents, u as mutateRunClaim } from "./record-events-D93DkpAx.js";
+import { Ct as readCommittedJsonLines, Ri as detachedSnapshot, Sr as isFiniteNumber, ai as authoredProfileDigest, bt as isNoEntError, i as RunParkedError, k as FileSpawnJournal, mt as loadSpawnForest, qi as assertValidBudget, r as assertResumeLimits, s as readRunParking, xr as failedItems, yt as appendCommittedJsonLine, zi as contentAddress } from "./supervisor-DDy0aEtR.js";
 import { f as RuntimeRunStateError, m as ValidationError } from "./errors-DodWX-cb.js";
 import { W as addSpend, X as errorMessage, ft as zeroSpend, q as cloneSpend } from "./sandbox-profile-compat-_fEtHPCj.js";
 import { a as withPursuitContext, t as composeRuntimeHooks } from "./runtime-hooks-tXpAarhW.js";
 import { a as applyExactAgentProfileDiff } from "./profile-6OLpbjvV.js";
-import { a as superviseRootProfile, ot as readRootStream, st as readRootStreamReceipt, t as supervise } from "./supervise-BuVVC7Bl.js";
-import { a as runCancelRequestFile, o as runCancellationFile, v as supervisorRunDir, x as workerCancellationsDir } from "./run-cancellation-CEDiqfhC.js";
-import { i as superviseAgentGraph } from "./graph-kgzTaNMA.js";
+import { a as superviseRootProfile, ot as readRootStream, st as readRootStreamReceipt, t as supervise } from "./supervise-BUOeThuB.js";
+import { a as runCancelRequestFile, o as runCancellationFile, v as supervisorRunDir, x as workerCancellationsDir } from "./run-cancellation-CzFhQGMa.js";
+import { i as superviseAgentGraph } from "./graph-CE_pzU31.js";
 import { agentProfileSchema, canonicalAgentProfileDigest, canonicalCandidateDigest, canonicalCandidateJson, sha256Bytes } from "@tangle-network/agent-interface";
 import { isRecord } from "@tangle-network/agent-core";
 import { createHash } from "node:crypto";
-import { existsSync, mkdirSync, readFileSync, readdirSync, renameSync } from "node:fs";
+import { closeSync, existsSync, fstatSync, mkdirSync, openSync, readFileSync, readdirSync, renameSync, unlinkSync } from "node:fs";
 import { basename, dirname, isAbsolute, join, relative, resolve, sep } from "node:path";
-import { access, mkdir, open, readFile, realpath, rm, rmdir, unlink } from "node:fs/promises";
+import { access, readFile, realpath } from "node:fs/promises";
 import { SearchRecorder, developmentClaim, incumbent, openSearchLedger, parseSearchLedgerLine, runSearch, uniform } from "@tangle-network/agent-eval/campaign";
 import { pathToFileURL } from "node:url";
 import { setTimeout as setTimeout$1 } from "node:timers/promises";
@@ -197,6 +197,7 @@ function deriveExecutionId(input) {
 var FileObserverJournal = class {
 	path;
 	pursuitId;
+	writeContext = captureRunWriteContext();
 	tail = Promise.resolve();
 	initialized = false;
 	sequence = 0;
@@ -240,7 +241,7 @@ var FileObserverJournal = class {
 			return Promise.reject(error);
 		}
 		let result;
-		const operation = this.tail.then(async () => {
+		const operation = this.tail.then(() => this.writeContext(async () => {
 			this.assertComplete();
 			try {
 				await this.initialize();
@@ -273,7 +274,7 @@ var FileObserverJournal = class {
 			this.sequence = record.sequence;
 			this.previousDigest = record.digest;
 			result = record;
-		});
+		}));
 		this.tail = operation.then(() => void 0, () => void 0);
 		return operation.then(() => {
 			if (!result) throw new Error("FileObserverJournal: append completed without a record");
@@ -2052,7 +2053,7 @@ var RunDirectoryLockedError = class extends Error {
 async function acquireRunDirectoryLock(runDir, runId, now = Date.now) {
 	const dir = resolve(runDir);
 	const path = resolve(dir, RUN_DIRECTORY_LOCK_FILE);
-	await mkdir(dir, { recursive: true });
+	mkdirSync(dir, { recursive: true });
 	const { pid, ...identity } = readLockHolderIdentity(process.pid);
 	const owner = {
 		pid,
@@ -2060,14 +2061,25 @@ async function acquireRunDirectoryLock(runDir, runId, now = Date.now) {
 		runId,
 		...identity
 	};
-	return withMutationGuard(path, async () => {
-		const existing = await readLockFile(path);
+	return mutateRunClaim(path, () => {
+		const existing = readLockFile(path);
 		if (existing.state === "holder" && !holderIsDead(existing)) throw new RunDirectoryLockedError(path, existing.holder);
-		if (existing.state !== "missing") await removeIfPresent(path);
+		let previousGeneration = 0;
+		try {
+			previousGeneration = Number(readFileSync(runGenerationPath(dir), "utf8"));
+			if (!Number.isSafeInteger(previousGeneration) || previousGeneration < 1) throw new Error("invalid run directory generation counter");
+		} catch (error) {
+			if (!isNoEntError(error)) throw error;
+		}
+		const generation = Math.max(previousGeneration, existing.state === "holder" ? existing.holder.generation ?? 0 : 0) + 1;
+		if (!Number.isSafeInteger(generation)) throw new Error("run directory generation exhausted");
+		writeAtomicDurableFile(runGenerationPath(dir), `${generation}\n`);
+		if (existing.state !== "missing") removeIfPresent(path);
 		const written = {};
 		const published = publishExclusiveDurableFile(path, (file) => {
 			written.holder = Object.freeze({
 				...owner,
+				generation,
 				lock: lockIdentity(file)
 			});
 			return `${JSON.stringify(written.holder)}\n`;
@@ -2077,7 +2089,16 @@ async function acquireRunDirectoryLock(runDir, runId, now = Date.now) {
 		let released = false;
 		return Object.freeze({
 			...holder,
+			generation,
 			path,
+			run: (operation) => {
+				if (holder.lock === void 0) throw new Error("run lock has no inode identity");
+				return runWithWriteClaim({
+					path,
+					generation,
+					lock: holder.lock
+				}, operation);
+			},
 			release: async () => {
 				if (released) return;
 				await releaseHolder(path, holder);
@@ -2114,16 +2135,16 @@ async function runDirectoryHolderIsLive(runDir) {
 		holder: existing.holder
 	};
 }
-async function readLockFile(path) {
+function readLockFile(path) {
 	let text;
 	let current;
 	try {
-		const file = await open(path, "r");
+		const file = openSync(path, "r");
 		try {
-			current = lockIdentity(await file.stat({ bigint: true }));
-			text = await file.readFile("utf8");
+			current = lockIdentity(fstatSync(file, { bigint: true }));
+			text = readFileSync(file, "utf8");
 		} finally {
-			await file.close();
+			closeSync(file);
 		}
 	} catch (error) {
 		if (isNoEntError(error)) return { state: "missing" };
@@ -2137,7 +2158,7 @@ async function readLockFile(path) {
 		throw new Error(`supervisePursuit: ${path} is not a readable lock file; remove it by hand`, { cause });
 	}
 	const record = parsed;
-	if (typeof record !== "object" || record === null || typeof record.pid !== "number" || !Number.isInteger(record.pid) || record.pid <= 0 || typeof record.startedAt !== "string" || typeof record.runId !== "string" || record.processStart !== void 0 && typeof record.processStart !== "string") throw new Error(`supervisePursuit: ${path} is not a readable lock file; remove it by hand`);
+	if (typeof record !== "object" || record === null || typeof record.pid !== "number" || !Number.isInteger(record.pid) || record.pid <= 0 || typeof record.startedAt !== "string" || typeof record.runId !== "string" || record.processStart !== void 0 && typeof record.processStart !== "string" || record.generation !== void 0 && (!Number.isSafeInteger(record.generation) || record.generation < 1)) throw new Error(`supervisePursuit: ${path} is not a readable lock file; remove it by hand`);
 	const { pid, startedAt, runId, processStart } = record;
 	const verdictRecord = parseLockHolderRecord(parsed);
 	const { pid: _pid, ...identity } = verdictRecord ?? { pid };
@@ -2147,6 +2168,7 @@ async function readLockFile(path) {
 			pid,
 			startedAt,
 			runId,
+			...record.generation === void 0 ? {} : { generation: record.generation },
 			...processStart === void 0 ? {} : { processStart },
 			...identity
 		}),
@@ -2160,36 +2182,18 @@ function holderIsDead(existing) {
 }
 /** Remove the lock only while it still names this holder; a reclaimed lock belongs to its new owner. */
 async function releaseHolder(path, holder) {
-	await withMutationGuard(path, async () => {
-		const existing = await readLockFile(path);
+	const deadline = performance.now() + 1e3;
+	while (existsSync(`${path}.guard`) && performance.now() < deadline) await setTimeout$1(10);
+	mutateRunClaim(path, () => {
+		const existing = readLockFile(path);
 		if (existing.state !== "holder") return;
-		if (existing.holder.pid !== holder.pid || existing.holder.startedAt !== holder.startedAt || existing.holder.runId !== holder.runId || existing.holder.processStart !== holder.processStart) return;
-		await removeIfPresent(path);
-	}, true);
-}
-/** Serialize stale reclamation and release; an abandoned guard requires operator recovery. */
-async function withMutationGuard(path, mutate, waitForRelease = false) {
-	const guard = `${path}.guard`;
-	const deadline = performance.now() + (waitForRelease ? 1e3 : 0);
-	for (;;) try {
-		await mkdir(guard);
-		break;
-	} catch (cause) {
-		if (waitForRelease && cause.code === "EEXIST" && performance.now() < deadline) {
-			await setTimeout$1(10);
-			continue;
-		}
-		throw new Error(`supervisePursuit: cannot mutate ${path}; inspect ${guard} and remove it only after confirming no lock mutation is active`, { cause });
-	}
-	try {
-		return await mutate();
-	} finally {
-		await rmdir(guard);
-	}
+		if (existing.holder.pid !== holder.pid || existing.holder.startedAt !== holder.startedAt || existing.holder.runId !== holder.runId || existing.holder.processStart !== holder.processStart || existing.holder.generation !== holder.generation || existing.holder.lock?.dev !== holder.lock?.dev || existing.holder.lock?.ino !== holder.lock?.ino || existing.holder.lock?.birthtimeNs !== holder.lock?.birthtimeNs) return;
+		removeIfPresent(path);
+	});
 }
-async function removeIfPresent(path) {
+function removeIfPresent(path) {
 	try {
-		await unlink(path);
+		unlinkSync(path);
 	} catch (error) {
 		if (!isNoEntError(error)) throw error;
 	}
@@ -2373,15 +2377,17 @@ async function reopenSettledRun(input) {
 			archivedAs
 		} }
 	});
-	setCancellationAside(runDir, operationId);
-	await rm(settlePath);
-	syncDurableDirectory(runDir);
+	fenceRunWrite(() => {
+		setCancellationAside(runDir, operationId);
+		unlinkSync(settlePath);
+		syncDurableDirectory(runDir);
+	});
 }
 /** The next free `settles/<n>.json`, or the archive that already holds these bytes (a reopen that
 *  crashed after archiving and before its journal row). */
 async function archiveName(runDir, settled) {
 	const dir = resolve(runDir, SETTLE_ARCHIVE_DIR);
-	mkdirSync(dir, { recursive: true });
+	fenceRunWrite(() => mkdirSync(dir, { recursive: true }));
 	const numbers = readdirSync(dir).map((name) => /^(\d+)\.json$/u.exec(name)?.[1]).filter((n) => n !== void 0).map(Number).sort((a, b) => a - b);
 	for (const n of numbers) if (await readBytes(join(dir, `${n}.json`)) === settled) return `${SETTLE_ARCHIVE_DIR}/${n}.json`;
 	return `${SETTLE_ARCHIVE_DIR}/${(numbers.at(-1) ?? 0) + 1}.json`;
@@ -2523,70 +2529,76 @@ async function runDurablePursuit(opts, execute, settled, reopen) {
 	const runId = opts.runId ?? "supervise";
 	const now = opts.now ?? Date.now;
 	const lock = await acquireRunDirectoryLock(runDir, runId, now);
-	let delivery;
-	let terminalState = "failed";
-	try {
-		const observer = createFileObserverHooks(observerPath, pursuitId);
-		if (reopen !== void 0) await reopen(observer.journal);
-		else {
-			const prior = await readSettleRecord(runDir);
-			if (prior !== void 0) throw new SettledRunDirectoryError(settlePath, prior.tree.root, runId);
-		}
-		await observer.journal.appendEvent(rootEvent(pursuitId, runId, "before", now()));
-		delivery = opts.observerDelivery === void 0 ? void 0 : startPursuitObserverDelivery(opts.observerDelivery, async () => projectPursuit(await observer.journal.read()));
-		let result;
+	return lock.run(async () => {
+		let delivery;
+		let terminalState = "failed";
 		try {
-			result = await execute({
-				hooks: withPursuitContext(pursuitId, composeRuntimeHooks(observer.hooks, opts.hooks)),
-				steerDir: opts.steerDir ?? supervisorRunDir(runDir, runId)
-			});
-		} catch (error) {
-			let pursuit;
-			let observerError;
-			try {
-				await observer.journal.appendEvent(rootEvent(pursuitId, runId, "error", now(), {
-					status: "failed",
-					error: errorMessage(error)
-				}));
-				pursuit = projectPursuit(await observer.journal.read());
-			} catch (failure) {
-				observerError = failure;
+			const observer = createFileObserverHooks(observerPath, pursuitId);
+			if (reopen !== void 0) await reopen(observer.journal);
+			else {
+				const prior = await readSettleRecord(runDir);
+				if (prior !== void 0) throw new SettledRunDirectoryError(settlePath, prior.tree.root, runId);
 			}
-			if (observerError !== void 0 || pursuit === void 0) throw new Error("supervisePursuit: Runtime failed and durable observer completeness could not be proven", { cause: new AggregateError(observerError === void 0 ? [error] : [error, observerError]) });
+			await observer.journal.appendEvent(rootEvent(pursuitId, runId, "before", now()));
+			delivery = opts.observerDelivery === void 0 ? void 0 : startPursuitObserverDelivery(opts.observerDelivery, async () => projectPursuit(await observer.journal.read()));
+			let result;
 			try {
-				const rootStream = await readRootStreamReceipt(runDir);
-				await writeFailureRecord(runDir, {
-					runId,
-					pursuitId,
-					at: new Date(now()).toISOString(),
-					error: {
-						name: errorName(error),
-						message: errorMessage(error)
-					},
-					...rootStream === void 0 ? {} : { rootStream }
+				result = await execute({
+					hooks: withPursuitContext(pursuitId, composeRuntimeHooks(observer.hooks, opts.hooks)),
+					steerDir: opts.steerDir ?? supervisorRunDir(runDir, runId)
 				});
-			} catch (recordError) {
-				throw new Error(`supervisePursuit: Runtime failed and the failure record ${failurePath} could not be written`, { cause: new AggregateError([error, recordError]) });
+			} catch (error) {
+				if (error instanceof RunParkedError && error.runId === runId) {
+					terminalState = "running";
+					throw error;
+				}
+				let pursuit;
+				let observerError;
+				try {
+					await observer.journal.appendEvent(rootEvent(pursuitId, runId, "error", now(), {
+						status: "failed",
+						error: errorMessage(error)
+					}));
+					pursuit = projectPursuit(await observer.journal.read());
+				} catch (failure) {
+					observerError = failure;
+				}
+				if (observerError !== void 0 || pursuit === void 0) throw new Error("supervisePursuit: Runtime failed and durable observer completeness could not be proven", { cause: new AggregateError(observerError === void 0 ? [error] : [error, observerError]) });
+				try {
+					const rootStream = await readRootStreamReceipt(runDir);
+					await writeFailureRecord(runDir, {
+						runId,
+						pursuitId,
+						at: new Date(now()).toISOString(),
+						error: {
+							name: errorName(error),
+							message: errorMessage(error)
+						},
+						...rootStream === void 0 ? {} : { rootStream }
+					});
+				} catch (recordError) {
+					throw new Error(`supervisePursuit: Runtime failed and the failure record ${failurePath} could not be written`, { cause: new AggregateError([error, recordError]) });
+				}
+				throw new SupervisePursuitError(error, pursuit, observerPath, failurePath);
 			}
-			throw new SupervisePursuitError(error, pursuit, observerPath, failurePath);
-		}
-		try {
-			await writeSettleRecord(runDir, settled(result));
-		} catch (cause) {
-			throw new Error(`supervisePursuit: Runtime settled but the settle record ${settlePath} could not be written`, { cause });
+			try {
+				await writeSettleRecord(runDir, settled(result));
+			} catch (cause) {
+				throw new Error(`supervisePursuit: Runtime settled but the settle record ${settlePath} could not be written`, { cause });
+			}
+			await observer.journal.appendEvent(rootEvent(pursuitId, runId, "after", now(), { status: "done" }));
+			terminalState = "done";
+			return Object.freeze({
+				result,
+				pursuit: projectPursuit(await observer.journal.read()),
+				observerPath,
+				settlePath
+			});
+		} finally {
+			await delivery?.close(terminalState);
+			await lock.release();
 		}
-		await observer.journal.appendEvent(rootEvent(pursuitId, runId, "after", now(), { status: "done" }));
-		terminalState = "done";
-		return Object.freeze({
-			result,
-			pursuit: projectPursuit(await observer.journal.read()),
-			observerPath,
-			settlePath
-		});
-	} finally {
-		await delivery?.close(terminalState);
-		await lock.release();
-	}
+	});
 }
 function rootEvent(pursuitId, runId, phase, timestamp, payload) {
 	return Object.freeze({
@@ -2629,12 +2641,12 @@ async function discoverDurableSupervisionRun(runDir) {
 			rootsBegunAt.set(record.root, typeof record.at === "string" ? record.at : null);
 			continue;
 		}
-		if (record.kind !== "event") continue;
-		const event = record.event;
-		if (!isRecord(event) || event.kind !== "spawned") continue;
-		if (!Object.hasOwn(event, "ownedTreeRoot")) continue;
-		if (typeof event.ownedTreeRoot !== "string" || event.ownedTreeRoot.length === 0) throw new Error(`${spawnJournalPath}: spawned event ownedTreeRoot must be a non-empty string when present`);
-		nestedRoots.add(event.ownedTreeRoot);
+		for (const event of spawnJournalRecordEvents(record)) {
+			if (!isRecord(event) || event.kind !== "spawned") continue;
+			if (!Object.hasOwn(event, "ownedTreeRoot")) continue;
+			if (typeof event.ownedTreeRoot !== "string" || event.ownedTreeRoot.length === 0) throw new Error(`${spawnJournalPath}: spawned event ownedTreeRoot must be a non-empty string when present`);
+			nestedRoots.add(event.ownedTreeRoot);
+		}
 	}
 	const streams = /* @__PURE__ */ new Map();
 	for await (const record of readCommittedJsonLines(coordinationLogPath, { allowMissing: true })) {
@@ -2670,6 +2682,6 @@ function compareText(left, right) {
 	return left < right ? -1 : left > right ? 1 : 0;
 }
 //#endregion
-export { FileObserverJournal, RUN_FORK_CORRELATION_KEYS, RunDirectoryLockedError, SettledRunDirectoryError, SupervisePursuitError, acquireRunDirectoryLock, assertPursuitVersions, deriveExecutionId, discoverDurableSupervisionRun, handleChatTurn, observerRecordDigest, projectPursuit, readFailureRecord, readRootStream, readRootStreamReceipt, readRunDirectoryLock, readSettleRecord, runDirectoryHolderIsLive, superviseGraphPursuit, supervisePursuit };
+export { FileObserverJournal, RUN_FORK_CORRELATION_KEYS, RunDirectoryLockedError, RunParkedError, SettledRunDirectoryError, SupervisePursuitError, acquireRunDirectoryLock, assertPursuitVersions, deriveExecutionId, discoverDurableSupervisionRun, handleChatTurn, observerRecordDigest, projectPursuit, readFailureRecord, readRootStream, readRootStreamReceipt, readRunDirectoryLock, readRunParking, readSettleRecord, runDirectoryHolderIsLive, superviseGraphPursuit, supervisePursuit };
 
 //# sourceMappingURL=durable.js.map