{
  "target": "https://github.com/pingdotgg/t3code/pull/17589",
  "base": "1451e7cb23034d29774338608368a15d8845df36",
  "lower": "26da5b009beb84d9db53ba93acdd5027afb61149",
  "upper": "1673308d37160f2478ba70a683fa06b8eb759976",
  "proofKind": "Headless Linux focused test output and isolated authenticated server WebSocket proof; no screenshots or live external-provider claim",
  "scopeCounts": {
    "mainCodex": 169,
    "finalCodex": 179,
    "mainLaunchOnPinnedMain": 79,
    "finalLaunch": 86,
    "mainMobilePermissions": 15,
    "finalMobilePermissions": 16
  },
  "results": [
    {
      "label": "Earlier integrated server proof, before type-neutral r4 changes",
      "tests": 417,
      "passed": 417,
      "failed": 0,
      "pending": 0,
      "success": true,
      "failures": []
    },
    {
      "label": "Final upper server scope",
      "tests": 101,
      "passed": 101,
      "failed": 0,
      "pending": 0,
      "success": true,
      "failures": []
    },
    {
      "label": "Final Codex scope",
      "tests": 179,
      "passed": 179,
      "failed": 0,
      "pending": 0,
      "success": true,
      "failures": []
    },
    {
      "label": "Final recovery scope",
      "tests": 97,
      "passed": 97,
      "failed": 0,
      "pending": 0,
      "success": true,
      "failures": []
    },
    {
      "label": "Mobile regression before fix",
      "tests": 16,
      "passed": 15,
      "failed": 1,
      "pending": 0,
      "success": false,
      "failures": [
        "queued worktree creation preserves the complete launch payload when retrying after a lost response and remount"
      ]
    },
    {
      "label": "Final mobile regression suite",
      "tests": 16,
      "passed": 16,
      "failed": 0,
      "pending": 0,
      "success": true,
      "failures": []
    },
    {
      "label": "Final related mobile suites",
      "tests": 138,
      "passed": 138,
      "failed": 0,
      "pending": 0,
      "success": true,
      "failures": []
    },
    {
      "label": "Repeated recovery and launch boundary regressions before fixes",
      "tests": 259,
      "passed": 255,
      "failed": 4,
      "pending": 0,
      "success": false,
      "failures": [
        "keeps the first launch intent on a same-key retry: crash after create before initial messa\u2026",
        "keeps the first launch intent on a same-key retry: accepted launch without an initial mess\u2026",
        "CodexAdapterV2 post-settle continuation delivers queued steering to Codex after recovery: two recovered siblings after the first \u2026",
        "CodexAdapterV2 post-settle continuation delivers queued steering to Codex after recovery: two recovered siblings with the second \u2026"
      ]
    },
    {
      "label": "Interrupted-sibling regression before fix",
      "tests": 179,
      "passed": 178,
      "failed": 1,
      "pending": 0,
      "success": false,
      "failures": [
        "CodexAdapterV2 post-settle continuation delivers queued steering to Codex after recovery: user-interrupted sibling before the sec\u2026"
      ]
    },
    {
      "label": "Current upper scope after stale-worktree fix",
      "tests": 111,
      "passed": 111,
      "failed": 0,
      "filteredOrPending": 0,
      "success": true
    },
    {
      "label": "Six Claude background-steering cases, focused timeout diagnosis",
      "tests": 286,
      "passed": 6,
      "failed": 0,
      "filteredOrPending": 280,
      "success": true
    },
    {
      "label": "Three isolated WorkflowOpeningBlock Chromium cases, CI diagnosis",
      "tests": 3,
      "passed": 3,
      "failed": 0,
      "filteredOrPending": 0,
      "success": true
    },
    {
      "label": "Fresh reuse workspace regressions before fix",
      "tests": 85,
      "passed": 83,
      "failed": 2,
      "success": false,
      "failure": "Root and new-worktree requests both ran preparation and checkpoint scope in the stale checkout."
    },
    {
      "label": "Current upper scope after atomic fresh-reuse checkout guard",
      "tests": 112,
      "passed": 112,
      "failed": 0,
      "pending": 0,
      "success": true
    },
    {
      "label": "Concurrent checkout claim regression before fix",
      "tests": 86,
      "passed": 85,
      "failed": 1,
      "success": false,
      "failure": "A second client binding after validation was admitted as the launch workspace; actual setup and checkpoint ran in that checkout."
    },
    {
      "label": "Round 9 current upper after test discriminant fix",
      "tests": 112,
      "passed": 112,
      "failed": 0,
      "pending": 0,
      "success": true
    }
  ],
  "live": {
    "head": "789d0f1a99f39868303296e5e9bf357189ef4a63",
    "base": "1451e7cb23034d29774338608368a15d8845df36",
    "transport": "actual authenticated WebSocket through T3 Browser",
    "scope": "no-initial-message launch, identical retry, changed-title conflict, persisted projection re-read",
    "providerStarted": false,
    "migration65": "applied successfully during isolated dev server startup",
    "results": {
      "first": {
        "tag": "Success",
        "threadId": "thread:project:4813c2ac-ef9f-472c-a44d-ff0edc0baccb:cb130834-d521-4ecb-922f-7a3de882dbb0",
        "resumed": false,
        "title": "Launch retry verification r4",
        "messages": 0,
        "runs": 0
      },
      "identical": {
        "tag": "Success",
        "threadId": "thread:project:4813c2ac-ef9f-472c-a44d-ff0edc0baccb:cb130834-d521-4ecb-922f-7a3de882dbb0",
        "resumed": true,
        "title": "Launch retry verification r4",
        "messages": 0,
        "runs": 0
      },
      "changed": {
        "tag": "Failure",
        "cause": [
          {
            "_tag": "Fail",
            "error": {
              "_tag": "OrchestrationV2ThreadLaunchError",
              "commandId": "command:composer-chat-17-live-r4-launch-1",
              "projectId": "4813c2ac-ef9f-472c-a44d-ff0edc0baccb",
              "message": "Failed to launch thread",
              "cause": {
                "name": "ThreadLaunchError",
                "message": "Thread launch command:composer-chat-17-live-r4-launch-1 failed during validate-intent.",
                "cause": "Launch request identity changed."
              }
            }
          }
        ]
      },
      "afterChanged": {
        "tag": "Success",
        "threadId": "thread:project:4813c2ac-ef9f-472c-a44d-ff0edc0baccb:cb130834-d521-4ecb-922f-7a3de882dbb0",
        "title": "Launch retry verification r4",
        "messages": 0,
        "runs": 0
      }
    },
    "limitations": [
      "No live external provider",
      "No real HTTP MCP credential rotation",
      "No physical OS restart",
      "No Electron/native client",
      "No synthetic-continuation visual proof"
    ]
  },
  "liveRuntimeApplicability": "The prior live proof covers new-thread launch identity and changed-title rejection. Later server change replaces stale workspace bindings only on fresh reuse; that behavior is proved separately by service/orchestrator/SQLite preparation tests, not a refreshed live checkout.",
  "typechecks": {
    "server": "passed on current upper 1673308d37160f2478ba70a683fa06b8eb759976; zero errors, zero warnings, 155 Effect suggestions",
    "contracts": "passed before type-neutral r4 changes",
    "mobile": "passed after r6 mobile payload change"
  },
  "lintAndFormatting": "Affected scopes passed; existing warnings documented in PR body/handoffs; no check bypassed",
  "reproductionCommands": [
    "nice -n 10 env RAYON_NUM_THREADS=2 node_modules/.bin/vp test run apps/server/src/orchestration-v2/Adapters/CodexAdapterV2.test.ts --maxWorkers=2",
    "nice -n 10 env RAYON_NUM_THREADS=2 node_modules/.bin/vp test run apps/server/src/orchestration-v2/ThreadLaunchService.test.ts apps/server/src/mcp/toolkits/project/handlers.test.ts packages/provider-core/src/server/orchestrationInstructions.test.ts apps/server/src/mcp/McpSessionRegistry.test.ts --maxWorkers=2",
    "nice -n 10 env RAYON_NUM_THREADS=2 node_modules/.bin/vp test run apps/mobile/src/state/use-thread-outbox-drain.permissions.test.ts --maxWorkers=2",
    "nice -n 10 env RAYON_NUM_THREADS=2 node_modules/.bin/tsc-rs --noEmit -p apps/server/tsconfig.json",
    "nice -n 10 env RAYON_NUM_THREADS=2 node_modules/.bin/tsc-rs --noEmit -p packages/contracts/tsconfig.json",
    "nice -n 10 env RAYON_NUM_THREADS=2 node_modules/.bin/tsc-rs --noEmit -p apps/mobile/tsconfig.json",
    "cd apps/web && nice -n 10 env RAYON_NUM_THREADS=2 ../../node_modules/.bin/vp test run src/components/chat/WorkflowOpeningBlock.browser.test.tsx --project browser --maxWorkers=2",
    "nice -n 10 env RAYON_NUM_THREADS=2 node_modules/.bin/vp test run apps/server/src/orchestration-v2/Adapters/ClaudeAdapterV2.test.ts -t \"moves the root.s foreground subagents to the background before a steer\" --maxWorkers=2"
  ],
  "limits": [
    "No live external-provider execution",
    "No real HTTP MCP credential rotation",
    "No physical OS restart",
    "No macOS, Windows, Electron, iOS or Android runtime proof",
    "No native reconnect timing or synthetic-continuation visual proof",
    "Later scoped passing runs are separate, not a fresh combined server run"
  ],
  "CIUpdate": "Lower actual jobs passed. Upper test-only narrowing fixes published TS2339; new head awaits fresh CI. Both layers require human approval. Prior upper web highlighting failure passed in isolated Chromium diagnosis.",
  "reviews": "Nine independent full-diff rounds; both final round-9 reviews report no confirmed source blocker. Human approval and current-head CI remain separate gates."
}