agent-inspect 6.29.1 → 6.29.3

package/CHANGELOG.md

@@ -1,5 +1,26 @@
 # Changelog
 
+## 6.29.3
+
+### Patch Changes
+
+- cf87bc9: Corrective patch after 6.29.2:
+
+  - Preserve MCP client prototype methods and private state via Proxy in `wrapMcpClient` (#420).
+  - Preserve `cacheWrite` and `reasoning` when converting persisted token usage back to TraceEvent metadata (#423 / #424).
+
+## 6.29.2
+
+### Patch Changes
+
+- 555052d: Published runtime and fidelity patches for the post-6.29.1 adoption train:
+
+  - Externalize `agent-inspect` from `@agent-inspect/mcp` so packed ESM/CJS consumers share the application `inspectRun` context (#413).
+  - Refuse `suite init` overwrite of an existing suite config (#411 / #412).
+  - Recognize nested manual `attributes.metadata.arguments|input|toolArguments` in tool-argument checks (#414).
+  - Persist bounded safe thrown error codes from manual instrumentation (#415).
+  - Normalize OpenAI Agents cached tokens from `input_tokens_details.cached_tokens` (#416).
+
 ## 6.29.1
 
 ### Patch Changes

package/README.md

@@ -212,7 +212,7 @@ The root package is enough for custom capture, the CLI, checks, and Evidence wor
 
 ## Status and documentation
 
-**Current published baseline:** **6.29.1** · persisted schema `1.0` · Node.js `>=20` · MIT.
+**Current published baseline:** **6.29.3** · persisted schema `1.0` · Node.js `>=20` · MIT.
 
 Legacy v0.1 and v0.2 traces remain readable. Check the npm badge and [changelog](CHANGELOG.md) for the current published version.
 

package/docs/API.md

@@ -474,6 +474,10 @@ Public methods:
 
 `capture` is explicit and metadata-only. `onSuccess: "metadata-only"` records safe type/length/key-count summaries in `outputSummary`; `onError: "metadata-only"` records thrown-value type/name summaries. It does not store raw return values, prompts, outputs, or thrown objects. Use `"none"` to disable a capture side.
 
+Persisted step/run errors keep `name` and `message`. When present, a **bounded safe** `code` is also retained: finite numeric `code`/`status` values (for example HTTP `404`) or short strings matching letters, digits, underscore, hyphen, dot, and colon (max 128 characters). Response bodies, headers, multiline values, and arbitrary enumerable error properties are not persisted.
+
+Pass structured tool arguments for contracts via step `metadata.arguments` / `metadata.input` / `metadata.toolArguments` (or top-level attributes with the same keys). TraceContract tool-argument checks read those locations; they do not enable default raw capture.
+
 `traceDir` and `silent` on `createInspector()` are context metadata for compatibility with existing helpers. They do not configure persistence or terminal output. Prefer writer-owned output configuration such as `fileWriter({ dir })` or `fileWriter({ filePath })`.
 
 `createInspectorRuntime()` is available from `agent-inspect/advanced` as the low-level isolation primitive. Most users should prefer `createInspector()` and `inspector.getDiagnostics()`.

package/docs/CLI.md

@@ -979,6 +979,8 @@ Options (shared):
 
 - `--template <name>` — PM/QA template (v5.4+): `customer-support-agent`, `refund-agent`, `sales-assistant`, `browser-task-agent`, `mcp-tool-agent`, `workflow-agent`, `rag-answer-agent`, `human-approval-agent`
 
+`suite init` creates a new `agent-inspect.suite.json` and **refuses to overwrite** an existing file (valid, malformed, or empty). Edit or delete the existing config explicitly before recreating it. There is no `--force` flag.
+
 `run` also supports:
 
 - `-o, --output <dir>` — write JSON run artifact (default: `.agent-inspect/suite-runs`)

package/docs/LIMITATIONS.md

@@ -9,6 +9,7 @@ This document states what AgentInspect **does not** provide today. It complement
 - **No vendor upload pipeline**: no built-in Langfuse/Braintrust/New Relic/Datadog direct exporters as live sinks.
 - **No automatic universal instrumentation** of every framework: integration is explicit (manual traces, log ingest, optional adapters).
 - **No retry execution engine:** TraceContract `retry` rules (including additive `retry.operations[]` recovery oracles in 6.27) evaluate attempt identity and evidence only. AgentInspect does not retry tools, remediates nothing, and does not treat a client `idempotencyKey` as proof of exactly-once writes. Write-tool timeout/unknown completion is unevaluable without authoritative idempotency evidence; write-retry is not safe by default.
+- **Transport facts are source-boundary only:** HTTP status, server-advertised retry delay, selected delay, and their sources must be captured at the transport/fetch boundary. An MCP wrapper cannot recover facts already collapsed into an SDK error message.
 - **Omitted-payload digests** (`createOmittedPayloadCommitment`) are bounded (1 MiB preflight). Digests prove omitted bytes existed; they are not redaction or authorization.
 
 ## Correlation metadata (v1.3.0)

package/docs/NETWORK-BEHAVIOR.md

@@ -27,6 +27,14 @@ AgentInspect is **local-first**. Core tracing writes local files and does **not*
 4. Prefer **redact / verify-safe** before any share or ingest path.
 5. Redaction is best-effort, not certification.
 
+## Transport evidence (HTTP / fetch / MCP)
+
+Capture **HTTP status**, **server-advertised retry delay**, **selected delay**, and their **sources** at the transport or fetch boundary that observes them.
+
+An MCP client wrapper or SDK error string cannot recover facts already collapsed into a message such as `"fetch failed"` or `"429 Too Many Requests"`. If the application needs TraceFacts or contracts over status/retry behavior, instrument the fetch/transport layer (or an observer that reads response headers before they are discarded).
+
+AgentInspect does not invent transport facts from error message text.
+
 ## Enabling Studio ingest
 
 See [SELF-HOSTING.md](./SELF-HOSTING.md) and [`@agent-inspect/studio`](https://github.com/rajudandigam/agent-inspect/tree/main/packages/studio). HTTP ingest remains off until you configure token and binding.

package/docs/TRACE-CONTRACTS.md

@@ -229,11 +229,22 @@ defineTraceContract({
 
 Missing structured argument evidence fails closed (`AI_CHECK_TOOL_ARGUMENT_EVIDENCE_UNAVAILABLE`). Findings never include full actual inputs.
 
+Manual instrumentation stores caller metadata under `attributes.metadata`. Tool-argument checks therefore also accept structured object/array evidence at:
+
+```text
+attributes.metadata.arguments
+attributes.metadata.input
+attributes.metadata.toolArguments
+```
+
+Precedence: top-level `attributes.arguments|input|toolArguments`, then nested metadata keys, then structured `inputSummary`. Preview strings are never parsed as JSON. This does not enable default raw argument capture.
+
 ### Capture capability matrix (tool-argument evidence)
 
 | Source | Structured input | Preview only | Digest only | Unavailable |
 | --- | :---: | :---: | :---: | :---: |
 | Manual `attributes.arguments` / `attributes.input` (object) | yes | — | — | — |
+| Manual `attributes.metadata.arguments` / `input` / `toolArguments` (object/array) | yes | — | — | — |
 | Manual `inputSummary` string | — | yes | — | for pointer checks |
 | AI SDK / LangChain metadata-only default | — | sometimes | — | typical |
 | OpenAI Agents metadata-only | — | sometimes | — | typical |

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "agent-inspect",
-  "version": "6.29.1",
+  "version": "6.29.3",
   "license": "MIT",
   "type": "module",
   "description": "Local evidence debugger and trajectory-test toolkit for TypeScript AI agents — execution trees, TraceContract checks, Evidence v2, and read-only MCP",
@@ -235,7 +235,7 @@
     "test:all": "pnpm run typecheck && pnpm run linked-versions:check && pnpm run build && pnpm run test && pnpm run size",
     "prepublish:checks": "pnpm run typecheck && pnpm run test && pnpm run test:coverage && pnpm run build && pnpm run fixtures:check && pnpm run recipes:check && pnpm run size && pnpm run linked-versions:check && pnpm run repo:health && pnpm run pack:smoke",
     "pack:dry-run": "pnpm run build && npm pack --dry-run",
-    "pack:smoke": "pnpm run build && node scripts/package-smoke.mjs && node scripts/packed-openai-agents-e2e.mjs && node scripts/packed-ai-sdk-e2e.mjs && node scripts/packed-quickstart-e2e.mjs && node scripts/packed-semantic-loop-e2e.mjs && node scripts/packed-swarm-loop-e2e.mjs && node scripts/evidence-ci-golden-paths.mjs",
+    "pack:smoke": "pnpm run build && node scripts/package-smoke.mjs && node scripts/packed-openai-agents-e2e.mjs && node scripts/packed-ai-sdk-e2e.mjs && node scripts/packed-mcp-e2e.mjs && node scripts/packed-quickstart-e2e.mjs && node scripts/packed-semantic-loop-e2e.mjs && node scripts/packed-swarm-loop-e2e.mjs && node scripts/evidence-ci-golden-paths.mjs",
     "linked-versions:check": "node scripts/check-linked-versions.mjs",
     "docs:commands": "node scripts/validate-doc-commands.mjs",
     "docs:links": "node scripts/validate-doc-links.mjs",

package/packages/cli/dist/{chunk-UK4HD5M3.mjs → chunk-TBTRH7KY.mjs}

@@ -386,7 +386,9 @@ function mapTokenUsageToMetadata(tokenUsage, attributes) {
       ...tokenUsage.input !== void 0 ? { input: tokenUsage.input } : {},
       ...tokenUsage.output !== void 0 ? { output: tokenUsage.output } : {},
       ...tokenUsage.total !== void 0 ? { total: tokenUsage.total } : {},
-      ...tokenUsage.cached !== void 0 ? { cached: tokenUsage.cached } : {}
+      ...tokenUsage.cached !== void 0 ? { cached: tokenUsage.cached } : {},
+      ...tokenUsage.cacheWrite !== void 0 ? { cacheWrite: tokenUsage.cacheWrite } : {},
+      ...tokenUsage.reasoning !== void 0 ? { reasoning: tokenUsage.reasoning } : {}
     };
   }
   return Object.keys(metadata).length > 0 ? metadata : void 0;
@@ -12486,5 +12488,5 @@ function renderGateReport(result, options = {}) {
 }
 
 export { COHORT_METRIC_IDS, DEFAULT_SUITE_ARTIFACTS_DIR, EVIDENCE_FORMAT_VERSION, EVIDENCE_HTML_FILENAME, EVIDENCE_MANIFEST_FILENAME, Redactor, TraceDirectory, TraceReadError, TreeBuilder, aggregateBundleSafeStatus, aggregateSessionCheckResults, analyzeCohort, applyProfileMetadataCaps, assertBundlePathContained, assertEvidenceRelativePath, buildActivitySummary, buildBundleMetadata, buildBundleSummaryMarkdown, buildEvidenceCausalFailureViewHtml, buildEvidenceCiPackage, buildEvidenceCircuitViewHtml, buildEvidenceContractsViewHtml, buildEvidenceDiffViewHtml, buildEvidenceHtmlShell, buildEvidenceManifest, buildEvidenceOutcomesViewHtml, buildEvidenceProvenanceViewHtml, buildEvidenceSafetyViewHtml, buildEvidenceTimelineViewHtml, buildEvidenceToolsLlmViewHtml, buildEvidenceTreeViewHtml, buildLocalExplanation, buildPlaceholderArtifact, buildRunSummary, buildRunTimeline, buildRunWhatSummary, buildSessionIndex, buildTraceStats, buildZipArchive, bundleFailsOnSafety, bundleRunAssetRelativePath, collectTraceSchemaVersions, compactAttributes, createBaselineRegressionRule, createLlmUsageRule, createMaxStepDurationRule, createObservedOutcomeRule, createRequireCompletedRule, createRunDepthRule, createRunDurationRule, createRunStatusRule, createSafetyOversizedAttributeRule, createSafetyRawContentRule, createSafetyRedactionRule, createSafetySecretPatternRule, createStallDetectionRule, createStructureCycleRule, createStructureOrphanRule, createStructureParallelWidthRule, createStructureRelationshipRule, createToolUsageRule, defaultBundleOutputPath, defaultSuiteConfigTemplate, diffRuns, diffTraceEvents, enrichSessionRunRecord, escapeHtml, escapeMarkdown, extractMetadata, extractOutcomesFromTraceEvents, filterMetasBySessionScope, filterTraces, flattenTree, formatDuration2 as formatDuration, formatStepLabel, formatTimestamp, gateHasThresholds, getIndent, getTraceFilePath, inferEvidenceFileRole, isAgentInspectTrace, isPersistedInspectEvent, loadSessionRunRecords, loadSuiteConfig, loadTraceMetadataList, manualTraceEventsToComparableRun, nanoid, normalizeBundleOutputPath, openTrace, parseCohortMetricList, parseDuration, parseDurationFilter, parseGateList, parseTraceJsonl, persistedInspectEventsToTraceEvents, renderActivitySummaryHuman, renderCohortReport, renderErrorLine, renderGateReport, renderObservedOutcomesHtml, renderObservedOutcomesMarkdown, renderRunDiff, renderRunWhat, renderStepLine, renderSuiteReport, renderTimeline, renderTraceStats, resolveBundleRunIds, resolveRedactionProfile, resolveSuiteTemplate, resolveTraceDir, runGate, runSuite, runTraceChecks, safeString, sanitizeBundleRunId, searchTraces, serializeEvidenceManifest, sha256Hex, stableJson, summarizeObservedOutcomes, summarizeSemanticParity, traceEventToPersistedInspectEvent, truncateName, truncateStringForProfile, validateEvent, validateSuiteConfig, verifyEvidenceDirectory, zeroKinds };
-//# sourceMappingURL=chunk-UK4HD5M3.mjs.map
-//# sourceMappingURL=chunk-UK4HD5M3.mjs.map
+//# sourceMappingURL=chunk-TBTRH7KY.mjs.map
+//# sourceMappingURL=chunk-TBTRH7KY.mjs.map