pi-session-mcp 0.8.3 → 0.12.0

package/docs/evaluation/live-rescreen-2026-10-17/approved-manifest.json

@@ -0,0 +1,315 @@
+{
+  "schema_version": "1.0.0",
+  "approved": true,
+  "server_commit": "406dae7e6b99aa5715960490e869bd3fc668cca2",
+  "screening_evidence": [],
+  "source_digest": "sha256:a4524626851448908b8d21ecbd7248c9956ce830406e81c45fda46c8eb6fffc5",
+  "tasks": [
+    {
+      "schema_version": "1.0.0",
+      "id": "T1",
+      "title": "Codebase explanation",
+      "family": "code-understanding",
+      "profiles": [
+        "P1"
+      ],
+      "views": [
+        "scout",
+        "docs-writer"
+      ],
+      "allowed_paths": [],
+      "prompt_hash": "sha256:d74a18bad9847c37741322c366c20ea05e6f2cc1bc046b164ae90ff2600c6000",
+      "fixture_hash": "sha256:579d5064e6529e3a08cbd7f3ecf9636f38ee138d17ee9b89d25fad81b8b69651",
+      "oracle_kind": "source-and-receipt",
+      "task_hash": "sha256:2b8a7d3cfbe985f25a9fe7e2c2e329e90f7c38a16f35ae3e180f1d9cd419531f"
+    },
+    {
+      "schema_version": "1.0.0",
+      "id": "T2",
+      "title": "Backend contract fix",
+      "family": "backend-fix",
+      "profiles": [
+        "P2",
+        "P3"
+      ],
+      "views": [
+        "implementer",
+        "reviewer"
+      ],
+      "allowed_paths": [
+        "contract.mjs"
+      ],
+      "prompt_hash": "sha256:3fb9a806d2c7de11f73c302088c5a3a1c48564f7588a7e392b6c98521840eccc",
+      "fixture_hash": "sha256:a5d8e952b1182e89398fff4fe0071d8537363aede0119f077ff4109f957a1b71",
+      "oracle_kind": "external-test-and-review",
+      "task_hash": "sha256:2a96d93af9503e98a474cecb240ecde2ee3052eaabef010fc16140a5b7123e7f"
+    },
+    {
+      "schema_version": "1.0.0",
+      "id": "T3",
+      "title": "Responsive accessible UI",
+      "family": "ui",
+      "profiles": [
+        "P4"
+      ],
+      "views": [
+        "ui-writer"
+      ],
+      "allowed_paths": [
+        "src/App.jsx",
+        "src/style.css"
+      ],
+      "prompt_hash": "sha256:cc72e58b0a7efcaf70170d339ac1197cd7e53fa862053107c6b9f4482323238c",
+      "fixture_hash": "sha256:d77cd778b5d4315636e7c18ea3ee92a260e1a4649ddf0e7a2b558b1e692553cd",
+      "oracle_kind": "browser-execution",
+      "task_hash": "sha256:399dbb8038164277fa7232300bb8e7e9a58fb0ba07beb08ff8dafe0febe08a6c"
+    },
+    {
+      "schema_version": "1.0.0",
+      "id": "T4",
+      "title": "Multi-module debugging",
+      "family": "debugging",
+      "profiles": [
+        "P2",
+        "P3"
+      ],
+      "views": [
+        "implementer",
+        "reviewer"
+      ],
+      "allowed_paths": [
+        "src/tax.mjs"
+      ],
+      "prompt_hash": "sha256:374611acced668d2e13e0bcd15eaf38749cb11888ea0791b4c1cd0878a1ab989",
+      "fixture_hash": "sha256:d98cb982a79a213408d181d1d30ba0051150c5762cf457f1cf9e01e5d14a037e",
+      "oracle_kind": "external-test-and-review",
+      "task_hash": "sha256:6b1a576b46aafea960738914848e68fb0499201f1345e3796579ab6a22a3e989"
+    },
+    {
+      "schema_version": "1.0.0",
+      "id": "T5",
+      "title": "Independent security review",
+      "family": "review",
+      "profiles": [
+        "P3"
+      ],
+      "views": [
+        "reviewer"
+      ],
+      "allowed_paths": [],
+      "prompt_hash": "sha256:97f8528b749b92838e0b3b4f4046a8fb81ffd9d14a60ecaeb33cf84c00a6858b",
+      "fixture_hash": "sha256:135a849ce7f8ceaec5a053466046b5e5d4f0d51227abca1038827e5ebb72a0cb",
+      "oracle_kind": "structured-findings",
+      "task_hash": "sha256:a76b843866bd20b4d40348575c7645e786947cc3eb49b0765b77f854c677e55e"
+    },
+    {
+      "schema_version": "1.0.0",
+      "id": "T6",
+      "title": "Regression test gap",
+      "family": "testing",
+      "profiles": [
+        "P4"
+      ],
+      "views": [
+        "tester"
+      ],
+      "allowed_paths": [],
+      "prompt_hash": "sha256:1d91df807efc91147ee7750d5730447d212de4ac9764f9a625a035e34fe82d68",
+      "fixture_hash": "sha256:e210625f68ee33354452e4ffa38d7a02759d7f1e480b6b60df9d668a5c3520b6",
+      "oracle_kind": "test-execution-and-coverage",
+      "task_hash": "sha256:8f15c34e5a1deffc17e754988e29aac62f1e28727985733019a728135b877433"
+    },
+    {
+      "schema_version": "1.0.0",
+      "id": "T7",
+      "title": "Infrastructure source review",
+      "family": "infrastructure-review",
+      "profiles": [
+        "P1"
+      ],
+      "views": [
+        "scout"
+      ],
+      "allowed_paths": [],
+      "prompt_hash": "sha256:cdacd6795b07e211e3a002a51aa49b2b1272e55cd5aec2aecc2b7e40851ef668",
+      "fixture_hash": "sha256:4053bd81269971ba8bae8b5bbc0066b43f50aaa8fcd5f638240fdd8b143a20f5",
+      "oracle_kind": "source-and-receipt",
+      "task_hash": "sha256:ef2252fc44f6750fd8630a0a48242b6ec602e5bfa7559226da6ec96aaf650fb3"
+    },
+    {
+      "schema_version": "1.0.0",
+      "id": "T8",
+      "title": "Evidence-based documentation draft",
+      "family": "documentation",
+      "profiles": [
+        "P1"
+      ],
+      "views": [
+        "docs-writer"
+      ],
+      "allowed_paths": [],
+      "prompt_hash": "sha256:2c35ae92d2aa1b62fdfe3c9bea0adb079d67ee9811db7c9a90971f0a84a87d0c",
+      "fixture_hash": "sha256:99aee6df8359ca777657bdc9fe6e5a09795e2b2a4d761c4919c5cb59ca950047",
+      "oracle_kind": "source-and-receipt",
+      "task_hash": "sha256:908d1af1652d71d8e5b917e16f656c62d443f97bae6000c7324c5a88763384a1"
+    },
+    {
+      "schema_version": "1.0.0",
+      "id": "T9",
+      "title": "Observed compaction continuation",
+      "family": "context-continuation",
+      "profiles": [
+        "P1",
+        "P2"
+      ],
+      "views": [
+        "scout",
+        "implementer"
+      ],
+      "allowed_paths": [],
+      "prompt_hash": "sha256:d8b100be1027e5eb54bd15581c3748fed50371dd5151165061cb53d9c0c4c709",
+      "fixture_hash": "sha256:c4d273343f9a6d324edfbf434fdacae1eb1bfc96a07cc08d992cf5ccead6aacb",
+      "oracle_kind": "observed-compaction-event",
+      "task_hash": "sha256:2567411f86cedee63fb7c601f786514351959ae2ac8e1485ef58775458e0d1b9"
+    }
+  ],
+  "runs": [
+    {
+      "run_id": "rescreen-1",
+      "task_id": "T1",
+      "task_hash": "sha256:2b8a7d3cfbe985f25a9fe7e2c2e329e90f7c38a16f35ae3e180f1d9cd419531f",
+      "profile_id": "P1",
+      "view": "scout",
+      "repetition": 1,
+      "stage": "screening",
+      "route": {
+        "provider": "openai-codex",
+        "model": "gpt-6-luna",
+        "effort": "low",
+        "auth_type": "oauth",
+        "billing_type": "subscription",
+        "data_path": "synthetic-fixture-to-selected-provider"
+      },
+      "permission_profile": "read-only",
+      "variant": "A",
+      "review_target": null,
+      "skills": [],
+      "extensions": []
+    },
+    {
+      "run_id": "rescreen-2",
+      "task_id": "T2",
+      "task_hash": "sha256:2a96d93af9503e98a474cecb240ecde2ee3052eaabef010fc16140a5b7123e7f",
+      "profile_id": "P2",
+      "view": "implementer",
+      "repetition": 1,
+      "stage": "screening",
+      "route": {
+        "provider": "openai-codex",
+        "model": "gpt-6-luna",
+        "effort": "low",
+        "auth_type": "oauth",
+        "billing_type": "subscription",
+        "data_path": "synthetic-fixture-to-selected-provider"
+      },
+      "permission_profile": "coding",
+      "variant": "A",
+      "review_target": null,
+      "skills": [],
+      "extensions": []
+    },
+    {
+      "run_id": "rescreen-3",
+      "task_id": "T1",
+      "task_hash": "sha256:2b8a7d3cfbe985f25a9fe7e2c2e329e90f7c38a16f35ae3e180f1d9cd419531f",
+      "profile_id": "P1",
+      "view": "scout",
+      "repetition": 1,
+      "stage": "screening",
+      "route": {
+        "provider": "openai-codex",
+        "model": "gpt-6.1-sol",
+        "effort": "medium",
+        "auth_type": "oauth",
+        "billing_type": "subscription",
+        "data_path": "synthetic-fixture-to-selected-provider"
+      },
+      "permission_profile": "read-only",
+      "variant": "A",
+      "review_target": null,
+      "skills": [],
+      "extensions": []
+    },
+    {
+      "run_id": "rescreen-4",
+      "task_id": "T2",
+      "task_hash": "sha256:2a96d93af9503e98a474cecb240ecde2ee3052eaabef010fc16140a5b7123e7f",
+      "profile_id": "P2",
+      "view": "implementer",
+      "repetition": 1,
+      "stage": "screening",
+      "route": {
+        "provider": "openai-codex",
+        "model": "gpt-6.1-sol",
+        "effort": "medium",
+        "auth_type": "oauth",
+        "billing_type": "subscription",
+        "data_path": "synthetic-fixture-to-selected-provider"
+      },
+      "permission_profile": "coding",
+      "variant": "A",
+      "review_target": null,
+      "skills": [],
+      "extensions": []
+    },
+    {
+      "run_id": "rescreen-5",
+      "task_id": "T5",
+      "task_hash": "sha256:a76b843866bd20b4d40348575c7645e786947cc3eb49b0765b77f854c677e55e",
+      "profile_id": "P3",
+      "view": "reviewer",
+      "repetition": 1,
+      "stage": "screening",
+      "route": {
+        "provider": "xai",
+        "model": "grok-4.7",
+        "effort": "high",
+        "auth_type": "oauth",
+        "billing_type": "subscription",
+        "data_path": "synthetic-fixture-to-selected-provider"
+      },
+      "permission_profile": "read-only",
+      "variant": "A",
+      "review_target": null,
+      "skills": [],
+      "extensions": []
+    },
+    {
+      "run_id": "rescreen-6",
+      "task_id": "T5",
+      "task_hash": "sha256:a76b843866bd20b4d40348575c7645e786947cc3eb49b0765b77f854c677e55e",
+      "profile_id": "P3",
+      "view": "reviewer",
+      "repetition": 2,
+      "stage": "screening",
+      "route": {
+        "provider": "xai",
+        "model": "grok-4.7",
+        "effort": "high",
+        "auth_type": "oauth",
+        "billing_type": "subscription",
+        "data_path": "synthetic-fixture-to-selected-provider"
+      },
+      "permission_profile": "read-only",
+      "variant": "A",
+      "review_target": null,
+      "skills": [],
+      "extensions": []
+    }
+  ],
+  "prerequisites": {
+    "baseline_passed": false,
+    "screening_passed": false,
+    "bottleneck": null
+  }
+}

package/docs/evaluation/live-rescreen-2026-10-17/audit-rescreen.mjs

@@ -0,0 +1,78 @@
+// Read-only live artifact validation; never starts an evaluation or provider.
+import assert from 'node:assert/strict';
+import {readFile,readdir,lstat,access,writeFile} from 'node:fs/promises';
+import {join} from 'node:path';
+import {tmpdir} from 'node:os';
+import {sourceBinding,hash,digest,snapshot} from '../../../.profile-eval/eval/profiles/files.js';
+import {manifestSchema,budgetSchema,resultSchema} from '../../../.profile-eval/eval/profiles/schemas.js';
+import {comparison} from '../../../.profile-eval/eval/profiles/comparison.js';
+const root=process.cwd();
+const raw='eval/results/live-rescreen-0e7491cd-68b7-400d-b738-489f9b9b6d67';
+const archive='docs/evaluation/live-rescreen-2026-10-17';
+const manifest=manifestSchema.parse(JSON.parse(await readFile(join(archive,'approved-manifest.json'),'utf8')));
+const budget=budgetSchema.parse(JSON.parse(await readFile(join(archive,'approved-budget.json'),'utf8')));
+const source=await sourceBinding(root);
+assert.equal(source.digest,manifest.source_digest);assert.equal(source.commit,manifest.server_commit);
+assert.equal(budget.manifest_digest,digest(manifest));assert.equal(manifest.runs.length,6);
+assert.equal(budget.max_starts,6);assert.equal(budget.parallelism,1);assert.equal(budget.max_retries,0);
+const entries=await readdir(join(raw,'runs'));
+const claims=entries.filter(n=>n.endsWith('.claim'));
+assert.ok(claims.length<=6);
+const results=[],artifacts=[],diagnostics=[];
+for(const run of manifest.runs){
+ const record=resultSchema.parse(JSON.parse(await readFile(join(raw,'runs',run.run_id+'.json'),'utf8')));
+ assert.equal(record.run_id,run.run_id);assert.equal(record.task_id,run.task_id);assert.equal(record.task_hash,run.task_hash);
+ assert.equal(record.stage,'screening');assert.equal(record.source_digest,manifest.source_digest);
+ assert.equal(record.server_commit,manifest.server_commit);assert.equal(record.repetition,run.repetition);
+ assert.equal(record.profile_id,run.profile_id);assert.equal(record.view,run.view);assert.equal(record.permission_profile,run.permission_profile);
+ assert.deepEqual(record.expected_diff_scope,run.permission_profile==='read-only'?[]:manifest.tasks.find(t=>t.id===run.task_id).allowed_paths);
+ assert.deepEqual(record.requested_route,run.route);assert.deepEqual(record.skills,[]);assert.deepEqual(record.extensions,[]);
+ assert.ok(record.observed_diff_scope.every(p=>record.expected_diff_scope.includes(p)));
+ assert.equal(record.profile_digest,digest({profile:run.profile_id,permission:run.permission_profile,route:run.route,skills:run.skills,extensions:run.extensions}));
+ assert.equal(record.fixture_revision,manifest.tasks.find(t=>t.id===run.task_id).fixture_hash);
+ assert.equal(record.cost_usd,null);assert.equal(record.usage,null);
+ if(record.started_at!==null){
+  assert.ok(claims.includes(run.run_id+'.claim'));
+  assert.equal(await readFile(join(raw,'runs',run.run_id+'.claim'),'utf8'),digest({run_id:run.run_id,source:manifest.source_digest}));
+  if(record.resolved_route!==null)assert.deepEqual(record.resolved_route,run.route);
+  assert.ok(record.duration_ms<=budget.run_timeout_ms+15000);
+  assert.equal(record.checks.find(c=>c.id==='observed-process-cleanup')?.status,'PASS');
+ }
+ if(record.harness_status==='PASS'){
+  assert.equal(record.checks.find(c=>c.id==='session-cleanup')?.status,'PASS');
+  assert.equal(record.checks.find(c=>c.id==='transport-route-binding')?.status,'PASS');
+ }
+ for(const artifact of record.artifacts){
+  const target=join(raw,'runs',artifact.path),st=await lstat(target);assert.ok(!st.isSymbolicLink());
+  const actual=st.isDirectory()?digest(await snapshot(target)):hash(await readFile(target));assert.equal(actual,artifact.hash);
+  artifacts.push({run_id:run.run_id,...artifact});
+ }
+ let retained=false;
+ if(record.started_at!==null){
+  assert.ok(record.scope_id.startsWith('pi-profile-eval-'));
+  try{await access(join(tmpdir(),record.scope_id));retained=true;}catch(e){assert.equal(e.code,'ENOENT');}
+ }else assert.equal(record.scope_id,'not-created');
+ if(record.task_status==='PASS'&&record.harness_status==='PASS')assert.equal(retained,false);
+ diagnostics.push({run_id:run.run_id,scope_id:record.scope_id,temporary_fixture_retained:retained});
+ results.push(record);
+}
+const started=results.filter(r=>r.started_at!==null),accepted=started.filter(r=>r.harness_status==='PASS'&&r.task_status==='PASS');
+assert.equal(started.length,claims.length);
+assert.equal(new Set(started.map(r=>r.scope_id)).size,started.length);
+for(let i=1;i<started.length;i++)assert.ok(Date.parse(started[i].started_at)>=Date.parse(started[i-1].ended_at));
+const first=started[0],last=started.at(-1);
+const interval=first&&last?Date.parse(last.ended_at)-Date.parse(first.started_at):0;
+assert.ok(interval<=budget.total_timeout_ms+15000);
+assert.deepEqual(JSON.parse(await readFile(join(raw,'runs/comparison.json'),'utf8')),comparison(results));
+const routes=[];
+for(const run of manifest.runs){
+ if(routes.some(r=>digest(r.route)===digest(run.route)))continue;
+ const matching=results.filter(r=>digest(r.requested_route)===digest(run.route));
+ const pass=matching.filter(r=>r.harness_status==='PASS'&&r.task_status==='PASS');
+ const review=run.profile_id==='P3';
+ const qualified=review?new Set(pass.filter(r=>r.task_id==='T5').map(r=>r.repetition)).size===2:['T1','T2'].every(t=>pass.some(r=>r.task_id===t));
+ routes.push({route:run.route,screening_gate:qualified?'PASS':'FAIL',eligible_for_separately_approved_pilot:qualified,evidence_run_ids:pass.map(r=>r.run_id),limitations:['Screening eligibility is not two-repeat role qualification or production approval.','Coding profiles have no OS file-scope boundary.']});
+}
+const audit={kind:'six-start-live-rescreen-audit',status:'PASS',source_binding:source,approved_manifest_digest:digest(manifest),claims:claims.length,started:started.length,accepted:accepted.length,rejected:started.filter(r=>r.task_status==='FAIL').length,blocked:results.filter(r=>r.harness_status==='BLOCKED').length,harness_passed:started.filter(r=>r.harness_status==='PASS').length,sequential_intervals_verified:true,unique_scopes_verified:true,artifacts,diagnostics,route_screening:routes,measured_interval_ms:interval,cost_usd:null,usage:null,cost_reason:'not-exposed-by-mcp',scope_violations:results.filter(r=>r.blockers.includes('observed_scope_violation')).map(r=>r.run_id),model_starts_beyond_approved_matrix:0,live_starts_used_across_authorized_waves:8+started.length,planning_envelope_remaining_if_30_total:30-8-started.length,limitations:['No raw model receipt archived; exact rejected response content cannot be reconstructed.','Cleanup covers observed sessions and process identities, not OS containment.','SDK-internal provider retries and provider cost/quota are not metered.']};
+await writeFile(join(archive,'audit.json'),JSON.stringify(audit,null,2)+'\n',{flag:'wx',mode:0o600});
+console.log(JSON.stringify({status:audit.status,started:audit.started,accepted:audit.accepted,harness_passed:audit.harness_passed,route_screening:routes.map(r=>({model:r.route.model,screening_gate:r.screening_gate}))}));

package/docs/evaluation/live-rescreen-2026-10-17/audit.json

@@ -0,0 +1,152 @@
+{
+  "kind": "six-start-live-rescreen-audit",
+  "status": "PASS",
+  "source_binding": {
+    "commit": "406dae7e6b99aa5715960490e869bd3fc668cca2",
+    "digest": "sha256:a4524626851448908b8d21ecbd7248c9956ce830406e81c45fda46c8eb6fffc5"
+  },
+  "approved_manifest_digest": "sha256:b3e80291c681ce8509ade8a387e278a65eae6eeb00cc709fbee76c5932206c97",
+  "claims": 5,
+  "started": 5,
+  "accepted": 3,
+  "rejected": 2,
+  "blocked": 1,
+  "harness_passed": 5,
+  "sequential_intervals_verified": true,
+  "unique_scopes_verified": true,
+  "artifacts": [
+    {
+      "run_id": "rescreen-1",
+      "path": "rescreen-1.diagnostic.json",
+      "hash": "sha256:c95e8462f49c8e96b9551dad55d33c2791805a3014c49d81672f1e0f3ab20de3"
+    },
+    {
+      "run_id": "rescreen-2",
+      "path": "rescreen-2.snapshot",
+      "hash": "sha256:3cfc2fcee9b65f78fb19830bf59ba89af06e1a46410435e7e293c9d67b18ae43"
+    },
+    {
+      "run_id": "rescreen-2",
+      "path": "rescreen-2.patch",
+      "hash": "sha256:6e5fe041c47508e7c390bf6e580faf93243e658cb93b6398e7c0c655b4cdb60e"
+    },
+    {
+      "run_id": "rescreen-4",
+      "path": "rescreen-4.snapshot",
+      "hash": "sha256:3cfc2fcee9b65f78fb19830bf59ba89af06e1a46410435e7e293c9d67b18ae43"
+    },
+    {
+      "run_id": "rescreen-4",
+      "path": "rescreen-4.patch",
+      "hash": "sha256:6e5fe041c47508e7c390bf6e580faf93243e658cb93b6398e7c0c655b4cdb60e"
+    },
+    {
+      "run_id": "rescreen-5",
+      "path": "rescreen-5.diagnostic.json",
+      "hash": "sha256:e974eaa3470cdbec0e7e21198a5d83f7d0a309f8a3c97c44cda47139384880da"
+    }
+  ],
+  "diagnostics": [
+    {
+      "run_id": "rescreen-1",
+      "scope_id": "pi-profile-eval-hOKIEl",
+      "temporary_fixture_retained": true
+    },
+    {
+      "run_id": "rescreen-2",
+      "scope_id": "pi-profile-eval-FGWr1Q",
+      "temporary_fixture_retained": false
+    },
+    {
+      "run_id": "rescreen-3",
+      "scope_id": "pi-profile-eval-4bpOaQ",
+      "temporary_fixture_retained": false
+    },
+    {
+      "run_id": "rescreen-4",
+      "scope_id": "pi-profile-eval-YDjN2s",
+      "temporary_fixture_retained": false
+    },
+    {
+      "run_id": "rescreen-5",
+      "scope_id": "pi-profile-eval-uiWdlk",
+      "temporary_fixture_retained": true
+    },
+    {
+      "run_id": "rescreen-6",
+      "scope_id": "not-created",
+      "temporary_fixture_retained": false
+    }
+  ],
+  "route_screening": [
+    {
+      "route": {
+        "provider": "openai-codex",
+        "model": "gpt-6-luna",
+        "effort": "low",
+        "auth_type": "oauth",
+        "billing_type": "subscription",
+        "data_path": "synthetic-fixture-to-selected-provider"
+      },
+      "screening_gate": "FAIL",
+      "eligible_for_separately_approved_pilot": false,
+      "evidence_run_ids": [
+        "rescreen-2"
+      ],
+      "limitations": [
+        "Screening eligibility is not two-repeat role qualification or production approval.",
+        "Coding profiles have no OS file-scope boundary."
+      ]
+    },
+    {
+      "route": {
+        "provider": "openai-codex",
+        "model": "gpt-6.1-sol",
+        "effort": "medium",
+        "auth_type": "oauth",
+        "billing_type": "subscription",
+        "data_path": "synthetic-fixture-to-selected-provider"
+      },
+      "screening_gate": "PASS",
+      "eligible_for_separately_approved_pilot": true,
+      "evidence_run_ids": [
+        "rescreen-3",
+        "rescreen-4"
+      ],
+      "limitations": [
+        "Screening eligibility is not two-repeat role qualification or production approval.",
+        "Coding profiles have no OS file-scope boundary."
+      ]
+    },
+    {
+      "route": {
+        "provider": "xai",
+        "model": "grok-4.7",
+        "effort": "high",
+        "auth_type": "oauth",
+        "billing_type": "subscription",
+        "data_path": "synthetic-fixture-to-selected-provider"
+      },
+      "screening_gate": "FAIL",
+      "eligible_for_separately_approved_pilot": false,
+      "evidence_run_ids": [],
+      "limitations": [
+        "Screening eligibility is not two-repeat role qualification or production approval.",
+        "Coding profiles have no OS file-scope boundary."
+      ]
+    }
+  ],
+  "measured_interval_ms": 74424,
+  "cost_usd": null,
+  "usage": null,
+  "cost_reason": "not-exposed-by-mcp",
+  "scope_violations": [],
+  "model_starts_beyond_approved_matrix": 0,
+  "live_starts_used_across_authorized_waves": 13,
+  "planning_envelope_remaining_if_30_total": 17,
+  "limitations": [
+    "No raw model receipt archived; exact rejected response content cannot be reconstructed.",
+    "Cleanup covers observed sessions and process identities, not OS containment.",
+    "SDK-internal provider retries and provider cost/quota are not metered."
+  ]
+}

package/docs/evaluation/live-rescreen-2026-10-17/backend-independent-checks.json

@@ -0,0 +1,38 @@
+{
+  "kind": "controller-reviewed-patches-and-independent-backend-checks",
+  "status": "PASS",
+  "assertions": 14,
+  "model_calls": 0,
+  "rows": [
+    {
+      "run_id": "rescreen-2",
+      "status": "PASS",
+      "assertions": 7,
+      "cases": [
+        "undefined-default",
+        "positive-integer",
+        "invalid-default",
+        "negative-integer-default",
+        "negative-fraction-default",
+        "zero-preserved",
+        "positive-fraction-preserved"
+      ],
+      "snapshot_entry_hash": "sha256:7df782c1888f40c7a6965803dfb8dcd71ea3db90d72920135e87a428f7b623c8"
+    },
+    {
+      "run_id": "rescreen-4",
+      "status": "PASS",
+      "assertions": 7,
+      "cases": [
+        "undefined-default",
+        "positive-integer",
+        "invalid-default",
+        "negative-integer-default",
+        "negative-fraction-default",
+        "zero-preserved",
+        "positive-fraction-preserved"
+      ],
+      "snapshot_entry_hash": "sha256:7df782c1888f40c7a6965803dfb8dcd71ea3db90d72920135e87a428f7b623c8"
+    }
+  ]
+}

package/docs/evaluation/live-rescreen-2026-10-17/checks.json

@@ -0,0 +1,54 @@
+{
+  "kind": "live-rescreen-closure-checks",
+  "status": "PASS",
+  "source_binding": {
+    "commit": "406dae7e6b99aa5715960490e869bd3fc668cca2",
+    "digest": "sha256:a4524626851448908b8d21ecbd7248c9956ce830406e81c45fda46c8eb6fffc5"
+  },
+  "archive_artifact_hashes": "PASS",
+  "prior_evidence_integrity": [
+    {
+      "dossier": "live-screening-2026-10-17",
+      "files_verified": 35,
+      "status": "PASS"
+    },
+    {
+      "dossier": "extension-audit-2026-10-17",
+      "files_verified": 22,
+      "status": "PASS"
+    },
+    {
+      "dossier": "node24-followup-2026-10-17",
+      "files_verified": 46,
+      "status": "PASS"
+    },
+    {
+      "dossier": "client-reconnect-2026-10-17",
+      "files_verified": 5,
+      "status": "PASS"
+    },
+    {
+      "dossier": "plan-continuation-2026-10-17",
+      "files_verified": 12,
+      "status": "PASS"
+    },
+    {
+      "dossier": "permission-gates-2026-10-17",
+      "files_verified": 9,
+      "status": "PASS"
+    }
+  ],
+  "independent_review": "PASS",
+  "independent_backend_assertions": 14,
+  "runner_exit_code": 1,
+  "expected_task_rejection_exit": true,
+  "started": 5,
+  "accepted": 3,
+  "blocked_not_started": 1,
+  "additional_live_calls_during_verification": 0,
+  "controller_source_changed": false,
+  "production_profiles_edited": false,
+  "credentials_provisioned": false,
+  "cost_usd": null,
+  "usage": null
+}

package/docs/evaluation/live-rescreen-2026-10-17/execution.json

@@ -0,0 +1,19 @@
+{
+  "kind": "live-rescreen-cli-execution",
+  "node": "24.22.0",
+  "exit_code": 1,
+  "exit_reason": "Two started task receipts failed independent acceptance; the same-combination fail gate blocked the sixth planned slot.",
+  "planned": 6,
+  "claims": 5,
+  "started": 5,
+  "harness_pass": 5,
+  "task_accepted": 3,
+  "task_rejected": 2,
+  "not_started_blocked": 1,
+  "controller_retries": 0,
+  "pilot_starts": 0,
+  "ablation_starts": 0,
+  "cost_usd": null,
+  "usage": null,
+  "cost_reason": "not-exposed-by-mcp"
+}