@tangle-network/traces 0.9.3 → 0.9.5

package/dist/{chunk-BKGWYQR4.js → chunk-S4TEBT7Q.js}

@@ -4455,9 +4455,9 @@ function skillDocumentNames(s) {
   }
   return [...names];
 }
-function telemetryStatus(measured, unsupported, unknown) {
-  if (measured > 0 && unsupported === 0 && unknown === 0) return "measured";
-  if (measured > 0) return "partial";
+function telemetryStatus(measured2, unsupported, unknown) {
+  if (measured2 > 0 && unsupported === 0 && unknown === 0) return "measured";
+  if (measured2 > 0) return "partial";
   if (unsupported > 0 && unknown === 0) return "unsupported";
   return "unknown";
 }
@@ -4684,6 +4684,12 @@ import {
 } from "@tangle-network/agent-eval/analyst";
 import { OtlpFileTraceStore } from "@tangle-network/agent-eval/traces";
 var GENERATED_TRACE_FILE_CEILING = 512 * 1024 * 1024;
+function mergeCostProvenance(first, second, totalCostUsd) {
+  if (!first || !second || first.kind === "uncaptured" || second.kind === "uncaptured") {
+    return { kind: "uncaptured", usd: null };
+  }
+  return first.kind === "estimated" || second.kind === "estimated" ? { kind: "estimated", usd: totalCostUsd } : { kind: "observed", usd: totalCostUsd };
+}
 async function analyzeSpans(spans, opts = {}) {
   if (spans.length === 0) throw new Error("analyzeSpans: no spans to analyze");
   const otlpPath = await writeOtlpFile(spans, opts.otlpOutPath);
@@ -4716,7 +4722,13 @@ async function analyzeSpans(spans, opts = {}) {
     });
     result.findings.push(...agResult.findings);
     result.per_analyst.push(...agResult.per_analyst);
-    result.total_cost_usd += agResult.total_cost_usd;
+    const totalCostUsd = result.total_cost_usd + agResult.total_cost_usd;
+    result.total_cost_provenance = mergeCostProvenance(
+      result.total_cost_provenance,
+      agResult.total_cost_provenance,
+      totalCostUsd
+    );
+    result.total_cost_usd = totalCostUsd;
   }
   return { otlpPath, execution, result };
 }
@@ -4842,9 +4854,12 @@ function decimal(value, digits = 1) {
     maximumFractionDigits: digits
   });
 }
+function measured(value, render) {
+  return value === null ? "not captured" : render(value);
+}
 function distributionRow(label, distribution, unit = "") {
   if (distribution.n === 0) return `| ${label} | 0 | not captured | not captured | not captured | not captured | not captured |`;
-  const value = (number) => `${decimal(number)}${unit}`;
+  const value = (number) => measured(number, (captured) => `${decimal(captured)}${unit}`);
   return `| ${label} | ${count(distribution.n)} | ${value(distribution.min)} | ${value(distribution.mean)} | ${value(distribution.p50)} | ${value(distribution.p95)} | ${value(distribution.max)} |`;
 }
 function tokenRows(usage) {
@@ -4856,14 +4871,23 @@ function tokenRows(usage) {
     ["Cache write", usage.totals.cacheWrite, usage.cacheWrite]
   ];
   return rows.map(
-    ([label, total, values]) => `| ${label} | ${count(total)} | ${count(values.n)} | ${decimal(values.mean)} | ${decimal(values.p50)} | ${decimal(values.p95)} | ${count(values.max)} |`
+    ([label, total, values]) => `| ${label} | ${count(total)} | ${count(values.n)} | ${measured(values.mean, decimal)} | ${measured(values.p50, decimal)} | ${measured(values.p95, decimal)} | ${measured(values.max, count)} |`
   );
 }
+function analysisCost(result) {
+  const cost = `$${result.total_cost_usd.toFixed(4)}`;
+  if (result.total_cost_provenance?.kind === "observed") return `Analysis cost: ${cost} observed.`;
+  if (result.total_cost_provenance?.kind === "estimated") return `Analysis cost: ${cost} estimated.`;
+  return `Known analysis cost: ${cost}; additional cost was not captured.`;
+}
 function renderExecution(report) {
   const execution = report.execution;
   const lines = ["## execution facts", ""];
   lines.push(
-    `- **Sessions:** ${count(execution.durationMs.n)}  |  **Model-call runs:** ${count(execution.modelCalls.runs)}  |  **Model-call events:** ${count(execution.modelCalls.events)}  |  **Sessions with tool errors:** ${count(execution.failures.runs)}/${count(execution.failures.reportingRuns)} (${decimal(execution.failures.fraction * 100, 2)}%)`
+    `- **Sessions:** ${count(execution.durationMs.n)}  |  **Model-call runs:** ${count(execution.modelCalls.runs)}  |  **Model-call events:** ${count(execution.modelCalls.events)}  |  **Sessions with execution errors:** ${count(execution.executionErrors.runs)}/${count(execution.executionErrors.reportingRuns)} (${measured(execution.executionErrors.fraction, (fraction) => `${decimal(fraction * 100, 2)}%`)})`
+  );
+  lines.push(
+    `- **Terminal outcomes:** ${count(execution.terminalOutcomes.succeeded)} succeeded  |  ${count(execution.terminalOutcomes.failed)} failed  |  ${count(execution.terminalOutcomes.cancelled)} cancelled  |  ${count(execution.terminalOutcomes.incomplete)} incomplete  |  ${count(execution.terminalOutcomes.unknown)} unknown`
   );
   lines.push("- **Task quality:** not measured; these traces do not include comparable task outcome labels.");
   lines.push("");
@@ -4936,7 +4960,7 @@ function renderReport(result, meta) {
   lines.push("");
   const sessionSummary = meta.unassignedTraceCount ? `${meta.sessionCount} identified session(s) + ${meta.unassignedTraceCount} ${plural(meta.unassignedTraceCount, "trace")} without a single stable session identity` : `${meta.sessionCount} session(s)`;
   lines.push(
-    `${sessionSummary}, ${meta.spanCount} spans \u2192 **${findingSummary}** across ${result.per_analyst.length} analyst(s). Analysis cost: $${result.total_cost_usd.toFixed(4)}.`
+    `${sessionSummary}, ${meta.spanCount} spans \u2192 **${findingSummary}** across ${result.per_analyst.length} analyst(s). ${analysisCost(result)}`
   );
   lines.push("");
   if (meta.unassignedTraceCount) {
@@ -6516,7 +6540,7 @@ async function sessionMeta(item) {
   if (branch) meta[ATTR.GIT_BRANCH] = branch;
   return meta;
 }
-var msToNano = (iso2) => parseIsoToEpochMs(iso2) * 1e6;
+var msToNano = (epochMs) => (BigInt(epochMs) * 1000000n).toString();
 function toTraceSpanEvents(spans, rootMeta) {
   return spans.map((s) => {
     const attributes = {};
@@ -6527,14 +6551,15 @@ function toTraceSpanEvents(spans, rootMeta) {
     attributes[ATTR.SESSION_ID] = s.trace_id;
     attributes[ATTR.INGEST_SOURCE] = INGEST_SOURCE_CLI;
     if (s.parent_span_id === null) Object.assign(attributes, rootMeta);
-    const startNano = msToNano(s.start_time);
+    const startMs = parseIsoToEpochMs(s.start_time);
+    const endMs = Math.max(startMs, parseIsoToEpochMs(s.end_time));
     return {
       traceId: s.trace_id,
       spanId: s.span_id,
       ...s.parent_span_id ? { parentSpanId: s.parent_span_id } : {},
       name: s.name,
-      startTimeUnixNano: startNano,
-      endTimeUnixNano: Math.max(startNano, msToNano(s.end_time)),
+      startTimeUnixNano: msToNano(startMs),
+      endTimeUnixNano: msToNano(endMs),
       attributes,
       ...s.status.code === "UNSET" ? {} : { status: { code: s.status.code, ...s.status.message ? { message: s.status.message } : {} } }
     };

package/dist/cli.js

@@ -40,12 +40,20 @@ import {
   writePolicyEvidenceFile,
   writeSessionIndexFile,
   writeTraceEvidenceExportFile
-} from "./chunk-BKGWYQR4.js";
+} from "./chunk-S4TEBT7Q.js";
 
 // src/cli.ts
 import { readFileSync } from "fs";
 import { readFile, stat } from "fs/promises";
 import { basename, resolve } from "path";
+import {
+  analyzeSupervisorRun,
+  findSupervisorRunDirs,
+  isUnavailable,
+  renderSupervisorRollupMarkdown,
+  renderSupervisorRunMarkdown,
+  rollupSupervisorRuns
+} from "@tangle-network/agent-eval/supervisor-run";
 function packageVersion() {
   const pkg = JSON.parse(readFileSync(new URL("../package.json", import.meta.url), "utf8"));
   if (typeof pkg.version !== "string" || !pkg.version) throw new Error("package.json is missing version");
@@ -92,6 +100,9 @@ function parseArgs(argv) {
       case "--cwd":
         a.cwd = next();
         break;
+      case "--supervisor-run-dir":
+        a.supervisorRunDir = next();
+        break;
       case "--since":
         a.since = next();
         break;
@@ -453,6 +464,7 @@ async function loadExportAttributes(args) {
   return attributes;
 }
 async function cmdAnalyze(args) {
+  if (args.supervisorRunDir) return cmdAnalyzeSupervisorRun(args.supervisorRunDir, args);
   const result = await investigate(args, { loadDefaultConfig: false });
   if (args.out) {
     await saveReport(args.out, result.report);
@@ -463,6 +475,32 @@ async function cmdAnalyze(args) {
     console.log(result.report);
   }
 }
+async function cmdAnalyzeSupervisorRun(runDir, args) {
+  const nested = await findSupervisorRunDirs(runDir);
+  let markdown;
+  if (nested.length > 0) {
+    markdown = renderSupervisorRollupMarkdown(
+      rollupSupervisorRuns(await Promise.all(nested.map((dir) => analyzeSupervisorRun(dir)))),
+      `Supervisor rollup \u2014 ${runDir}`
+    );
+  } else {
+    const report = await analyzeSupervisorRun(runDir);
+    if (isUnavailable(report.orchestration.workersSpawned)) {
+      throw new Error(
+        `no supervisor run found at ${runDir} \u2014 expected <runDir>/ws/.loops/supervisor/<id>, or a parent directory containing such runs`
+      );
+    }
+    markdown = renderSupervisorRunMarkdown(report);
+  }
+  if (args.out) {
+    await saveReport(args.out, markdown);
+    console.log(
+      `supervisor report \u2192 ${args.out}  (${nested.length > 0 ? `${nested.length} runs` : "1 run"})`
+    );
+  } else {
+    console.log(markdown);
+  }
+}
 async function collectImportedSpans(args) {
   if (!args.input) throw new Error("input file missing");
   const attributes = await loadExportAttributes(args);
@@ -787,6 +825,7 @@ function usage() {
 Commands:
   list      List discovered sessions
   analyze   Run analyst suite + loop/waste pipelines over sessions or an input file
+            (--supervisor-run-dir <dir> reports a supervision tree instead)
   investigate Run typed investigation flow, including BYO config + evidence-backed actions
   improve   Write findings, evidence, report, and canonical trace artifacts
   convert   Emit OTLP-JSONL only (HALO: use analyze --analyzer halo)
@@ -804,6 +843,10 @@ Options:
   --last <n>       Most-recent N sessions
   --session <id|path> Analyze one listed session ID or one explicit harness session file
   --cwd <dir>      Filter sessions by working directory
+  --supervisor-run-dir <dir>
+                   analyze: report a SUPERVISION TREE instead of harness sessions \u2014
+                   steers, spawn waves, concurrency, idle wall, cost by role,
+                   accepted vs rejected. Rolls up when the dir holds many runs.
   --since <t>      upload: window, 30m / 2h / 7d or an ISO date (default 24h); analyze: ISO cutoff
   --out <path>     Write report to a file
   --dir <path>     improve: write artifacts to this directory

package/dist/index.js

@@ -117,7 +117,7 @@ import {
   writeSessionIndexFile,
   writeTraceEvidenceExportFile,
   writeTraceImprovementArtifacts
-} from "./chunk-BKGWYQR4.js";
+} from "./chunk-S4TEBT7Q.js";
 
 // src/collect.ts
 async function collectSessions(opts = {}) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@tangle-network/traces",
-  "version": "0.9.3",
+  "version": "0.9.5",
   "description": "Point it at your coding-agent session traces (Claude Code, Codex, OpenCode, Gemini, Pi, …) and get failure-mode + efficiency findings. CLI + SDK over the @tangle-network/agent-eval analyst suite — observe live sessions, run your own analysts, redact, and upload to the Tangle Intelligence Platform.",
   "type": "module",
   "license": "MIT",
@@ -45,7 +45,7 @@
   },
   "dependencies": {
     "@ax-llm/ax": "^23.0.1",
-    "@tangle-network/agent-eval": "^0.122.8"
+    "@tangle-network/agent-eval": "^0.133.0"
   },
   "devDependencies": {
     "@types/node": "^22.0.0",