@kontourai/flow-agents 3.2.0 → 3.4.0

package/evals/fixtures/learning-review-proposals/effect-follow-up/sessions/task-lr-ef-4/trust.bundle

@@ -0,0 +1,21 @@
+{
+  "schemaVersion": 3,
+  "source": "flow-agents/workflow-sidecar;statusFunctionVersion=1",
+  "claims": [
+    {
+      "id": "task-lr-ef-4-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass",
+      "subjectType": "workflow-check",
+      "subjectId": "task-lr-ef-4/unit-tests",
+      "facet": "flow-agents.workflow",
+      "claimType": "builder.verify.tests",
+      "fieldOrBehavior": "unit tests pass",
+      "value": "pass",
+      "createdAt": "2026-07-04T00:00:00Z",
+      "updatedAt": "2026-07-04T00:00:00Z",
+      "status": "verified"
+    }
+  ],
+  "evidence": [],
+  "events": [],
+  "policies": []
+}

package/evals/fixtures/learning-review-proposals/effect-follow-up/sessions/task-lr-ef-5/trust.bundle

@@ -0,0 +1,21 @@
+{
+  "schemaVersion": 3,
+  "source": "flow-agents/workflow-sidecar;statusFunctionVersion=1",
+  "claims": [
+    {
+      "id": "task-lr-ef-5-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass",
+      "subjectType": "workflow-check",
+      "subjectId": "task-lr-ef-5/unit-tests",
+      "facet": "flow-agents.workflow",
+      "claimType": "builder.verify.tests",
+      "fieldOrBehavior": "unit tests pass",
+      "value": "pass",
+      "createdAt": "2026-07-04T00:00:00Z",
+      "updatedAt": "2026-07-04T00:00:00Z",
+      "status": "verified"
+    }
+  ],
+  "evidence": [],
+  "events": [],
+  "policies": []
+}

package/evals/fixtures/learning-review-proposals/pattern-present/economics.jsonl

@@ -0,0 +1,6 @@
+{"schema":"kontour.console.economics","version":"0.1","run_id":"pp-r1","at":1751500000000,"task_slug":"task-lr-pp-1","model":"claude-opus-4-8","cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.1,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":1,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"pp-r2","at":1751500001000,"task_slug":"task-lr-pp-2","model":"claude-opus-4-8","cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.1,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":1,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"pp-r3","at":1751500002000,"task_slug":"task-lr-pp-3","model":"claude-opus-4-8","cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.1,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":1,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"pp-r4","at":1751500003000,"task_slug":"task-lr-pp-4","model":"claude-opus-4-8","cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.2,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":1,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"pp-r5","at":1751500004000,"task_slug":"task-lr-pp-5","model":"claude-opus-4-8","cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.2,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":1,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"pp-r6","at":1751500005000,"task_slug":"task-lr-pp-6","model":"claude-opus-4-8","cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.2,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":1,"verification_verdict":"PASS"}}

package/evals/fixtures/learning-review-proposals/pattern-present/expected-aggregates.json

@@ -0,0 +1,30 @@
+{
+  "by_kit": [
+    {
+      "kit_id": "builder",
+      "runs": 6,
+      "first_half_avg_cost_usd": 0.1,
+      "second_half_avg_cost_usd": 0.2,
+      "cost_trend_pct": 100,
+      "first_half_findings_total": 6,
+      "second_half_findings_total": 6,
+      "findings_delta_pct": 0,
+      "avg_wall_clock_s": 100,
+      "avg_human_wait_s": 10,
+      "route_back_rate": 0.5,
+      "caught_false_completions_total": 6
+    }
+  ],
+  "by_gate": [
+    {
+      "gate_id": "unit tests pass",
+      "fire_count": 4,
+      "correct_count": 1,
+      "false_block_count": 3,
+      "missed_block_count": 0,
+      "false_block_rate": 0.75,
+      "avg_wall_clock_s_when_fired": 100,
+      "avg_human_wait_s_when_fired": 10
+    }
+  ]
+}

package/evals/fixtures/learning-review-proposals/pattern-present/expected-aggregates.md

@@ -0,0 +1,66 @@
+# `pattern-present` — hand-computed aggregate arithmetic (AC1)
+
+Six `kontour.console.economics` records in `economics.jsonl`, ordered ascending by `.at`
+(`pp-r1` .. `pp-r6`), all joined (via `sessions/<task_slug>/trust.bundle`, one `builder.verify.tests`
+claim each) to `kit_id = "builder"`. Four of the six sessions (`task-lr-pp-1..4`) also carry a
+`gate-review.inquiries.json` with one `InquiryRecord` each for gate `"unit tests pass"`; the
+remaining two (`task-lr-pp-5`, `task-lr-pp-6`) have no gate-review join (counted in `notes`).
+
+Raw per-record inputs (identical across records except `cost` and — for the first four — the
+gate calibration):
+
+| run_id | at | cost.estimated_cost_usd | findings (high+low) | time.wall_clock_s | time.human_wait_s | iterations.count | iterations.route_backs | defects.caught_false_completions | gate calibration |
+| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
+| pp-r1 | 1751500000000 | 0.10 | 1+1=2 | 100 | 10 | 2 | 1 | 1 | correct |
+| pp-r2 | 1751500001000 | 0.10 | 1+1=2 | 100 | 10 | 2 | 1 | 1 | false_block |
+| pp-r3 | 1751500002000 | 0.10 | 1+1=2 | 100 | 10 | 2 | 1 | 1 | false_block |
+| pp-r4 | 1751500003000 | 0.20 | 1+1=2 | 100 | 10 | 2 | 1 | 1 | false_block |
+| pp-r5 | 1751500004000 | 0.20 | 1+1=2 | 100 | 10 | 2 | 1 | 1 | (no gate join) |
+| pp-r6 | 1751500005000 | 0.20 | 1+1=2 | 100 | 10 | 2 | 1 | 1 | (no gate join) |
+
+## `by_kit[]` — `kit_id: "builder"` (runs = 6)
+
+`half_point = floor(6 / 2) = 3` → first half = {pp-r1, pp-r2, pp-r3} (oldest 3 by `.at`), second
+half = {pp-r4, pp-r5, pp-r6} (newest 3).
+
+- `first_half_avg_cost_usd` = (0.10 + 0.10 + 0.10) / 3 = **0.10**
+- `second_half_avg_cost_usd` = (0.20 + 0.20 + 0.20) / 3 = **0.20**
+- `cost_trend_pct` = ((0.20 − 0.10) / 0.10) × 100 = **100** ( ≥ `LR_COST_RISE_PCT` default 25 )
+- `first_half_findings_total` = 2 + 2 + 2 = **6**
+- `second_half_findings_total` = 2 + 2 + 2 = **6**
+- `findings_delta_pct` = first half is non-zero (6), so `((6 − 6) / 6) × 100` = **0** ( ≤
+  `LR_FLAT_FINDINGS_PCT` default 10 — flat, not rising )
+- `avg_wall_clock_s` = (100 × 6) / 6 = **100** (all six records, not half-split)
+- `avg_human_wait_s` = (10 × 6) / 6 = **10**
+- `route_back_rate` = sum(route_backs) / sum(count) = (1×6) / (2×6) = 6 / 12 = **0.5**
+- `caught_false_completions_total` = 1 × 6 = **6**
+
+`runs (6) >= LR_MIN_KIT_SAMPLE (6)` AND `cost_trend_pct (100) >= LR_COST_RISE_PCT (25)` AND
+`findings_delta_pct (0) <= LR_FLAT_FINDINGS_PCT (10)` → **`kit-review-cost-inflation` fires** for
+`{kind:"kit", id:"builder"}` (AC2).
+
+## `by_gate[]` — `gate_id: "unit tests pass"` (fire_count = 4)
+
+Gate rows come only from `task-lr-pp-1..4`'s `gate-review.inquiries.json` (one row each);
+`task-lr-pp-5`/`-6` contribute none (2 sessions without a gate-review join, per `notes`).
+
+- `fire_count` = 4 (pp-r1 correct, pp-r2/pp-r3/pp-r4 false_block)
+- `correct_count` = 1 (pp-r1)
+- `false_block_count` = 3 (pp-r2, pp-r3, pp-r4)
+- `missed_block_count` = 0
+- `false_block_rate` = 3 / 4 = **0.75** ( ≥ `LR_GATE_FALSE_BLOCK_RATE` default 0.5 )
+- `avg_wall_clock_s_when_fired` = mean of the parent record's `time.wall_clock_s` over the 4
+  distinct records with a `fired:true` row for this gate (pp-r1..pp-r4, all 100) = **100**
+- `avg_human_wait_s_when_fired` = same, `time.human_wait_s` (all 10) = **10**
+
+`fire_count (4) >= LR_MIN_GATE_SAMPLE (3)` AND `false_block_rate (0.75) >= LR_GATE_FALSE_BLOCK_RATE
+(0.5)` → **`gate-false-block-review` fires** for `{kind:"gate", id:"unit tests pass"}` (AC2).
+`gate-well-calibrated` does NOT fire (`false_block_count` is 3, not 0).
+
+These are the exact numbers asserted (4-decimal-rounded per the contract's `round4`, though every
+value here is already exact at 0–2 decimal places) in
+`evals/integration/test_learning_review_proposals.sh` and recorded verbatim in
+`expected-aggregates.json` — computed from the formulas in
+`docs/specs/learning-review-proposals-contract.md`, independently of running
+`scripts/telemetry/learning-review-proposals.sh` (then cross-checked against the script's actual
+output as a confirmation, not a derivation).

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-1/gate-review.inquiries.json

@@ -0,0 +1,26 @@
+[
+  {
+    "id": "task-lr-pp-1-gr-1",
+    "inquiry": {
+      "id": "task-lr-pp-1-gr-1",
+      "question": "Was gate action on claim task-lr-pp-1/unit-tests justified?",
+      "askedBy": "gate-review",
+      "askedAt": "2026-07-03T00:05:00Z",
+      "target": { "subjectType": "workflow-check", "subjectId": "task-lr-pp-1/unit-tests", "fieldOrBehavior": "unit tests pass" }
+    },
+    "outcome": "matched",
+    "resolutionPath": { "claimIds": ["task-lr-pp-1-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass"] },
+    "answer": {
+      "status": "disputed",
+      "value": {
+        "calibration": "correct",
+        "advisoryFix": "No action needed; gate correctly blocked on a disputed claim.",
+        "gateFired": true,
+        "sessionSlug": "task-lr-pp-1"
+      }
+    },
+    "inputSnapshot": [{ "claimId": "task-lr-pp-1-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass", "status": "disputed" }],
+    "statusFunctionVersion": "1",
+    "resolvedAt": "2026-07-03T00:05:00Z"
+  }
+]

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-1/trust.bundle

@@ -0,0 +1,21 @@
+{
+  "schemaVersion": 3,
+  "source": "flow-agents/workflow-sidecar;statusFunctionVersion=1",
+  "claims": [
+    {
+      "id": "task-lr-pp-1-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass",
+      "subjectType": "workflow-check",
+      "subjectId": "task-lr-pp-1/unit-tests",
+      "facet": "flow-agents.workflow",
+      "claimType": "builder.verify.tests",
+      "fieldOrBehavior": "unit tests pass",
+      "value": "pass",
+      "createdAt": "2026-07-03T00:00:00Z",
+      "updatedAt": "2026-07-03T00:00:00Z",
+      "status": "disputed"
+    }
+  ],
+  "evidence": [],
+  "events": [],
+  "policies": []
+}

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-2/gate-review.inquiries.json

@@ -0,0 +1,26 @@
+[
+  {
+    "id": "task-lr-pp-2-gr-1",
+    "inquiry": {
+      "id": "task-lr-pp-2-gr-1",
+      "question": "Was gate action on claim task-lr-pp-2/unit-tests justified?",
+      "askedBy": "gate-review",
+      "askedAt": "2026-07-03T00:05:00Z",
+      "target": { "subjectType": "workflow-check", "subjectId": "task-lr-pp-2/unit-tests", "fieldOrBehavior": "unit tests pass" }
+    },
+    "outcome": "matched",
+    "resolutionPath": { "claimIds": ["task-lr-pp-2-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass"] },
+    "answer": {
+      "status": "verified",
+      "value": {
+        "calibration": "false_block",
+        "advisoryFix": "Investigate why the gate blocked when the claim was already verified passing.",
+        "gateFired": true,
+        "sessionSlug": "task-lr-pp-2"
+      }
+    },
+    "inputSnapshot": [{ "claimId": "task-lr-pp-2-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass", "status": "verified" }],
+    "statusFunctionVersion": "1",
+    "resolvedAt": "2026-07-03T00:05:00Z"
+  }
+]

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-2/trust.bundle

@@ -0,0 +1,21 @@
+{
+  "schemaVersion": 3,
+  "source": "flow-agents/workflow-sidecar;statusFunctionVersion=1",
+  "claims": [
+    {
+      "id": "task-lr-pp-2-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass",
+      "subjectType": "workflow-check",
+      "subjectId": "task-lr-pp-2/unit-tests",
+      "facet": "flow-agents.workflow",
+      "claimType": "builder.verify.tests",
+      "fieldOrBehavior": "unit tests pass",
+      "value": "pass",
+      "createdAt": "2026-07-03T00:00:00Z",
+      "updatedAt": "2026-07-03T00:00:00Z",
+      "status": "verified"
+    }
+  ],
+  "evidence": [],
+  "events": [],
+  "policies": []
+}

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-3/gate-review.inquiries.json

@@ -0,0 +1,26 @@
+[
+  {
+    "id": "task-lr-pp-3-gr-1",
+    "inquiry": {
+      "id": "task-lr-pp-3-gr-1",
+      "question": "Was gate action on claim task-lr-pp-3/unit-tests justified?",
+      "askedBy": "gate-review",
+      "askedAt": "2026-07-03T00:05:00Z",
+      "target": { "subjectType": "workflow-check", "subjectId": "task-lr-pp-3/unit-tests", "fieldOrBehavior": "unit tests pass" }
+    },
+    "outcome": "matched",
+    "resolutionPath": { "claimIds": ["task-lr-pp-3-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass"] },
+    "answer": {
+      "status": "verified",
+      "value": {
+        "calibration": "false_block",
+        "advisoryFix": "Investigate why the gate blocked when the claim was already verified passing.",
+        "gateFired": true,
+        "sessionSlug": "task-lr-pp-3"
+      }
+    },
+    "inputSnapshot": [{ "claimId": "task-lr-pp-3-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass", "status": "verified" }],
+    "statusFunctionVersion": "1",
+    "resolvedAt": "2026-07-03T00:05:00Z"
+  }
+]

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-3/trust.bundle

@@ -0,0 +1,21 @@
+{
+  "schemaVersion": 3,
+  "source": "flow-agents/workflow-sidecar;statusFunctionVersion=1",
+  "claims": [
+    {
+      "id": "task-lr-pp-3-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass",
+      "subjectType": "workflow-check",
+      "subjectId": "task-lr-pp-3/unit-tests",
+      "facet": "flow-agents.workflow",
+      "claimType": "builder.verify.tests",
+      "fieldOrBehavior": "unit tests pass",
+      "value": "pass",
+      "createdAt": "2026-07-03T00:00:00Z",
+      "updatedAt": "2026-07-03T00:00:00Z",
+      "status": "verified"
+    }
+  ],
+  "evidence": [],
+  "events": [],
+  "policies": []
+}

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-4/gate-review.inquiries.json

@@ -0,0 +1,26 @@
+[
+  {
+    "id": "task-lr-pp-4-gr-1",
+    "inquiry": {
+      "id": "task-lr-pp-4-gr-1",
+      "question": "Was gate action on claim task-lr-pp-4/unit-tests justified?",
+      "askedBy": "gate-review",
+      "askedAt": "2026-07-03T00:05:00Z",
+      "target": { "subjectType": "workflow-check", "subjectId": "task-lr-pp-4/unit-tests", "fieldOrBehavior": "unit tests pass" }
+    },
+    "outcome": "matched",
+    "resolutionPath": { "claimIds": ["task-lr-pp-4-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass"] },
+    "answer": {
+      "status": "verified",
+      "value": {
+        "calibration": "false_block",
+        "advisoryFix": "Investigate why the gate blocked when the claim was already verified passing.",
+        "gateFired": true,
+        "sessionSlug": "task-lr-pp-4"
+      }
+    },
+    "inputSnapshot": [{ "claimId": "task-lr-pp-4-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass", "status": "verified" }],
+    "statusFunctionVersion": "1",
+    "resolvedAt": "2026-07-03T00:05:00Z"
+  }
+]

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-4/trust.bundle

@@ -0,0 +1,21 @@
+{
+  "schemaVersion": 3,
+  "source": "flow-agents/workflow-sidecar;statusFunctionVersion=1",
+  "claims": [
+    {
+      "id": "task-lr-pp-4-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass",
+      "subjectType": "workflow-check",
+      "subjectId": "task-lr-pp-4/unit-tests",
+      "facet": "flow-agents.workflow",
+      "claimType": "builder.verify.tests",
+      "fieldOrBehavior": "unit tests pass",
+      "value": "pass",
+      "createdAt": "2026-07-03T00:00:00Z",
+      "updatedAt": "2026-07-03T00:00:00Z",
+      "status": "verified"
+    }
+  ],
+  "evidence": [],
+  "events": [],
+  "policies": []
+}

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-5/trust.bundle

@@ -0,0 +1,21 @@
+{
+  "schemaVersion": 3,
+  "source": "flow-agents/workflow-sidecar;statusFunctionVersion=1",
+  "claims": [
+    {
+      "id": "task-lr-pp-5-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass",
+      "subjectType": "workflow-check",
+      "subjectId": "task-lr-pp-5/unit-tests",
+      "facet": "flow-agents.workflow",
+      "claimType": "builder.verify.tests",
+      "fieldOrBehavior": "unit tests pass",
+      "value": "pass",
+      "createdAt": "2026-07-03T00:00:00Z",
+      "updatedAt": "2026-07-03T00:00:00Z",
+      "status": "verified"
+    }
+  ],
+  "evidence": [],
+  "events": [],
+  "policies": []
+}

package/evals/fixtures/learning-review-proposals/pattern-present/sessions/task-lr-pp-6/trust.bundle

@@ -0,0 +1,21 @@
+{
+  "schemaVersion": 3,
+  "source": "flow-agents/workflow-sidecar;statusFunctionVersion=1",
+  "claims": [
+    {
+      "id": "task-lr-pp-6-gate-claim-unit-tests.flow-agents-workflow.unit-tests-pass",
+      "subjectType": "workflow-check",
+      "subjectId": "task-lr-pp-6/unit-tests",
+      "facet": "flow-agents.workflow",
+      "claimType": "builder.verify.tests",
+      "fieldOrBehavior": "unit tests pass",
+      "value": "pass",
+      "createdAt": "2026-07-03T00:00:00Z",
+      "updatedAt": "2026-07-03T00:00:00Z",
+      "status": "verified"
+    }
+  ],
+  "evidence": [],
+  "events": [],
+  "policies": []
+}

package/evals/fixtures/learning-review-proposals/repeat-window/economics.jsonl

@@ -0,0 +1,6 @@
+{"schema":"kontour.console.economics","version":"0.1","run_id":"rw-r1","at":1751800000000,"task_slug":null,"cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.10,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":0,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"rw-r2","at":1751800001000,"task_slug":null,"cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.10,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":0,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"rw-r3","at":1751800002000,"task_slug":null,"cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.10,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":0,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"rw-r4","at":1751800003000,"task_slug":null,"cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.30,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":0,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"rw-r5","at":1751800004000,"task_slug":null,"cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.30,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":0,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"rw-r6","at":1751800005000,"task_slug":null,"cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.30,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":0,"verification_verdict":"PASS"}}

package/evals/fixtures/learning-review-proposals/under-threshold/economics.jsonl

@@ -0,0 +1,3 @@
+{"schema":"kontour.console.economics","version":"0.1","run_id":"ut-r1","at":1751700000000,"task_slug":null,"cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.10,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":0,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"ut-r2","at":1751700001000,"task_slug":null,"cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.20,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":0,"verification_verdict":"PASS"}}
+{"schema":"kontour.console.economics","version":"0.1","run_id":"ut-r3","at":1751700002000,"task_slug":null,"cost":{"input_tokens":1000,"output_tokens":500,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"estimated_cost_usd":0.30,"by_model":[]},"time":{"wall_clock_s":100,"human_wait_s":10},"iterations":{"count":2,"route_backs":1},"defects":{"gate_fires":1,"findings_by_severity":{"critical":0,"high":1,"medium":0,"low":1},"caught_false_completions":0,"verification_verdict":"PASS"}}

package/evals/fixtures/reconcile-preflight/disputed-critique-unsuperseded.json

@@ -0,0 +1,48 @@
+{
+  "schemaVersion": 5,
+  "source": "reconcile-preflight-fixture:disputed-critique-unsuperseded",
+  "note": "#356 shape 3: an un-superseded workflow.critique.review claim with a 'fail' verdict derives status 'disputed' — session-local-failed divergence (a disputed critique always blocks reconcile unless a later same-reviewer pass superseded it via metadata.superseded_by).",
+  "claims": [
+    {
+      "id": "c-disputed-critique",
+      "claimType": "workflow.critique.review",
+      "value": "fail",
+      "status": "disputed",
+      "subjectId": "reconcile-preflight-fixture/review-1",
+      "facet": "flow-agents.workflow",
+      "subjectType": "workflow-critique",
+      "fieldOrBehavior": "Found a critical security issue in the auth handler.",
+      "createdAt": "2026-07-04T00:00:00Z",
+      "updatedAt": "2026-07-04T00:00:00Z",
+      "impactLevel": "medium",
+      "verificationPolicyId": "policy:workflow.critique.review",
+      "metadata": { "origin": "critique", "reviewer": "tool-code-reviewer", "reviewed_at": "2026-07-04T00:00:00Z" }
+    }
+  ],
+  "evidence": [],
+  "policies": [
+    {
+      "id": "policy:workflow.critique.review",
+      "claimType": "workflow.critique.review",
+      "requiredEvidence": [],
+      "acceptanceCriteria": ["A verified verification event must support a workflow.critique.review claim."],
+      "reviewAuthority": "system",
+      "validityRule": { "kind": "manual" },
+      "stalenessTriggers": [],
+      "conflictRules": [],
+      "impactLevel": "medium"
+    }
+  ],
+  "events": [
+    {
+      "id": "evt-disputed-critique",
+      "claimId": "c-disputed-critique",
+      "status": "disputed",
+      "actor": "flow-agents/workflow-sidecar",
+      "method": "validation",
+      "evidenceIds": [],
+      "createdAt": "2026-07-04T00:00:00Z",
+      "verifiedAt": "2026-07-04T00:00:00Z"
+    }
+  ]
+}

package/evals/fixtures/reconcile-preflight/standalone-disputed-session-local.json

@@ -0,0 +1,59 @@
+{
+  "schemaVersion": 5,
+  "source": "reconcile-preflight-fixture:standalone-disputed-session-local",
+  "note": "#356 shape 4 (the #292/#384 variant): a NON-critique session-local claim (workflow.check.external, human_attestation-backed) with a re-derived status of 'disputed' and no waiver. Distinct from shape 3 (an un-superseded disputed CRITIQUE) only in claimType/subject — both land on the same session-local-failed divergence, proving a disjoint pre-existing failure can never be smuggled in as a standalone claim; the fix hint directs the agent to document it as prose in a waived non-command summary instead.",
+  "claims": [
+    {
+      "id": "c-standalone-disputed",
+      "claimType": "workflow.check.external",
+      "value": "fail",
+      "status": "disputed",
+      "subjectId": "reconcile-preflight-fixture/pre-existing-flaky-integration",
+      "facet": "flow-agents.workflow",
+      "subjectType": "workflow-check",
+      "fieldOrBehavior": "pre-existing integration test failure unrelated to this change",
+      "createdAt": "2026-07-04T00:00:00Z",
+      "updatedAt": "2026-07-04T00:00:00Z",
+      "impactLevel": "high",
+      "verificationPolicyId": "policy:workflow.check.external:human_attestation"
+    }
+  ],
+  "evidence": [
+    {
+      "id": "ev-standalone-disputed",
+      "claimId": "c-standalone-disputed",
+      "evidenceType": "human_attestation",
+      "method": "attestation",
+      "sourceRef": "reconcile-preflight-fixture/evidence.json",
+      "excerptOrSummary": "pre-existing integration test failure, unrelated to this change, observed but not fixed",
+      "observedAt": "2026-07-04T00:00:00Z",
+      "collectedBy": "flow-agents/workflow-sidecar",
+      "passing": false
+    }
+  ],
+  "policies": [
+    {
+      "id": "policy:workflow.check.external:human_attestation",
+      "claimType": "workflow.check.external",
+      "requiredEvidence": ["human_attestation"],
+      "acceptanceCriteria": ["A verified verification event must support a workflow.check.external claim."],
+      "reviewAuthority": "system",
+      "validityRule": { "kind": "manual" },
+      "stalenessTriggers": [],
+      "conflictRules": [],
+      "impactLevel": "high"
+    }
+  ],
+  "events": [
+    {
+      "id": "evt-standalone-disputed",
+      "claimId": "c-standalone-disputed",
+      "status": "disputed",
+      "actor": "flow-agents/workflow-sidecar",
+      "method": "attestation",
+      "evidenceIds": ["ev-standalone-disputed"],
+      "createdAt": "2026-07-04T00:00:00Z",
+      "verifiedAt": "2026-07-04T00:00:00Z"
+    }
+  ]
+}

package/evals/fixtures/telemetry/usage-transcript-sample.jsonl

@@ -0,0 +1,4 @@
+{"type":"user","message":{"role":"user","content":[{"type":"text","text":"hermetic fixture prompt"}]}}
+{"type":"assistant","message":{"model":"claude-opus-4-8","usage":{"input_tokens":1200,"output_tokens":3400,"cache_creation_input_tokens":500,"cache_read_input_tokens":600000}}}
+{"type":"assistant","message":{"model":"claude-opus-4-8","usage":{"input_tokens":300,"output_tokens":900,"cache_creation_input_tokens":0,"cache_read_input_tokens":0}}}
+{"type":"assistant","message":{"model":"claude-fable-5","usage":{"input_tokens":50,"output_tokens":200,"cache_creation_input_tokens":0,"cache_read_input_tokens":0}}}

package/evals/fixtures/trust-reconcile-exploits/mcp-degrade.json

@@ -0,0 +1,42 @@
+{
+  "schemaVersion": 5,
+  "source": "ws8-exploit-fixture:mcp-degrade",
+  "note": "ISSUE 492 DELIVERABLE 2: an MCP-shaped self-reported pass must not launder into a CI-verified kind:'command' check. The evidence deliberately carries a test_output execution.label that names an MCP tool, not a run-baseline.sh manifest command. CI must reject it as not-run / not-in-manifest; at most this kind of non-native pass can remain a session-local external attestation, never RECONCILED.",
+  "claims": [
+    {
+      "id": "c-mcp-degrade-test",
+      "claimType": "workflow.check.test",
+      "value": "pass",
+      "status": "verified",
+      "subjectId": "mcp/unit-tests",
+      "facet": "flow-agents.workflow",
+      "subjectType": "workflow-check",
+      "fieldOrBehavior": "unit tests passed via MCP runner",
+      "createdAt": "2026-07-07T00:00:00Z",
+      "updatedAt": "2026-07-07T00:00:00Z",
+      "impactLevel": "high",
+      "verificationPolicyId": "policy:workflow.check.test"
+    }
+  ],
+  "evidence": [
+    {
+      "id": "ev-mcp-degrade-test",
+      "claimId": "c-mcp-degrade-test",
+      "evidenceType": "test_output",
+      "method": "validation",
+      "sourceRef": "mcp://runner/run_tests",
+      "excerptOrSummary": "MCP runner reported tests passed",
+      "observedAt": "2026-07-07T00:00:00Z",
+      "collectedBy": "mcp__runner__run_tests",
+      "passing": true,
+      "execution": {
+        "runner": "mcp",
+        "label": "mcp__runner__run_tests",
+        "isError": false,
+        "exitCode": 0
+      }
+    }
+  ],
+  "policies": [],
+  "events": []
+}