refactor(deep-scan): compose ordinary scans and merge their results

.github/workflows/node-ci.yml

@@ -393,13 +393,17 @@ jobs:
         with:
           package_json_file: package.json
           cache: true
-          cache_dependency_path: plugins/codex-security/mcp-app/pnpm-lock.yaml
+          cache_dependency_path: |
+            sdk/typescript/pnpm-lock.yaml
+            plugins/codex-security/mcp-app/pnpm-lock.yaml
       - name: Set up Node.js
         uses: actions/setup-node@820762786026740c76f36085b0efc47a31fe5020 # v7.0.0
         with:
           node-version: "22.13.0"
       - name: Install dependencies
-        run: pnpm --dir plugins/codex-security/mcp-app install --frozen-lockfile
+        run: |
+          pnpm --dir sdk/typescript install --frozen-lockfile
+          pnpm --dir plugins/codex-security/mcp-app install --frozen-lockfile
       - name: Install ripgrep
         run: |
           sudo apt-get update
@@ -435,25 +439,35 @@ jobs:
       matrix:
         os: [ubuntu-latest, macos-latest, windows-latest]
     steps:
-      - name: Checkout plugin source without the SDK
+      - name: Checkout plugin and shared SDK source
         uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
         with:
           persist-credentials: false
-          sparse-checkout: plugins/codex-security
+          sparse-checkout: |
+            plugins/codex-security
+            sdk/typescript
       - name: Set up pnpm
         uses: pnpm/action-setup@ea17c68df8912ef543352723c149a84f56e3d413 # v6.1.0
         with:
           package_json_file: package.json
           cache: true
-          cache_dependency_path: plugins/codex-security/mcp-app/pnpm-lock.yaml
+          cache_dependency_path: |
+            plugins/codex-security/mcp-app/pnpm-lock.yaml
+            sdk/typescript/pnpm-lock.yaml
       - name: Set up Node.js
         uses: actions/setup-node@820762786026740c76f36085b0efc47a31fe5020 # v7.0.0
         with:
           node-version: "22.13.0"
-      - name: Install plugin dependencies
-        run: pnpm --dir plugins/codex-security/mcp-app install --frozen-lockfile
+      - name: Install plugin build dependencies
+        run: |
+          pnpm --dir sdk/typescript install --frozen-lockfile
+          pnpm --dir plugins/codex-security/mcp-app install --frozen-lockfile
       - name: Build and test the standalone host runtime
         run: node --test plugins/codex-security/mcp-app/scripts/test_host_build.mjs
+      - name: Test native executable discovery
+        run: node --test plugins/codex-security/mcp-app/tests/test_native_executable.mjs
+      - name: Test native scan preparation
+        run: node --test --test-name-pattern="native preparation requires" plugins/codex-security/mcp-app/tests/test_native_scan.mjs
 
   plugin-source:
     name: plugin source contracts
@@ -589,7 +603,6 @@ jobs:
         shell: pwsh
         run: ./sdk/typescript/scripts/prepare-windows-test-root.ps1
       - name: Test shard ${{ matrix.shard }}
-        timeout-minutes: 10
         env:
           TEMP: ${{ steps.windows-temp.outputs.path }}
           TMP: ${{ steps.windows-temp.outputs.path }}

.github/workflows/test-quality.yml

@@ -32,21 +32,45 @@ jobs:
     if: ${{ !cancelled() && (inputs.native-artifacts-ready || needs.native.result == 'success') }}
     name: ${{ matrix.os }} / ${{ matrix.mode }}
     runs-on: ${{ matrix.os }}
-    timeout-minutes: ${{ matrix.os == 'windows-latest' && (matrix.mode == 'baseline' || matrix.mode == 'isolated') && 45 || 30 }}
+    timeout-minutes: ${{ matrix.os == 'windows-latest' && (matrix.mode == 'baseline' && 60 || matrix.mode == 'parallel' && 45) || 30 }}
     env:
       CODEX_SECURITY_PROPERTY_SEED: ${{ github.event_name == 'pull_request' && 1 || github.run_number }}
     strategy:
       fail-fast: false
       matrix:
         os: [ubuntu-latest, windows-latest]
         mode: [baseline, isolated, parallel]
+        exclude:
+          - os: windows-latest
+            mode: isolated
         include:
           - mode: baseline
             args: ""
           - mode: isolated
             args: --isolate
           - mode: parallel
             args: --parallel=2
+          - os: windows-latest
+            mode: isolated-1
+            args: --isolate --shard=1/7
+          - os: windows-latest
+            mode: isolated-2
+            args: --isolate --shard=2/7
+          - os: windows-latest
+            mode: isolated-3
+            args: --isolate --shard=3/7
+          - os: windows-latest
+            mode: isolated-4
+            args: --isolate --shard=4/7
+          - os: windows-latest
+            mode: isolated-5
+            args: --isolate --shard=5/7
+          - os: windows-latest
+            mode: isolated-6
+            args: --isolate --shard=6/7
+          - os: windows-latest
+            mode: isolated-7
+            args: --isolate --shard=7/7
           - os: windows-latest
             mode: shard-1
             args: --shard=1/7
@@ -92,7 +116,7 @@ jobs:
             plugins/codex-security/mcp-app/pnpm-lock.yaml
       - uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
         with:
-          bun-version: "1.3.13"
+          bun-version: "1.4.2"
       - name: Install dependencies
         run: |
           pnpm --dir sdk/typescript install --frozen-lockfile
@@ -104,6 +128,7 @@ jobs:
         run: ./sdk/typescript/scripts/prepare-windows-test-root.ps1
       - name: Test runner mode
         env:
+          CODEX_SECURITY_TEST_TIMEOUT_MS: ${{ runner.os == 'Windows' && '120000' || '30000' }}
           TEMP: ${{ steps.windows-temp.outputs.path || runner.temp }}
           TMP: ${{ steps.windows-temp.outputs.path || runner.temp }}
           TMPDIR: ${{ steps.windows-temp.outputs.path || runner.temp }}
@@ -152,7 +177,7 @@ jobs:
           comparison_status=0
           for os in ubuntu-latest windows-latest; do
             for mode in isolated parallel; do
-              node sdk/typescript/scripts/compare-test-reports.mjs "reports/runner-$os-baseline.xml" "reports/runner-$os-$mode.xml" >> "$GITHUB_STEP_SUMMARY" || comparison_status=1
+              node sdk/typescript/scripts/compare-test-reports.mjs "reports/runner-$os-baseline.xml" "reports/runner-$os-$mode*.xml" >> "$GITHUB_STEP_SUMMARY" || comparison_status=1
             done
           done
           node sdk/typescript/scripts/compare-test-reports.mjs reports/runner-windows-latest-baseline.xml 'reports/runner-windows-latest-shard-*.xml' >> "$GITHUB_STEP_SUMMARY" || comparison_status=1

CONTRIBUTING.md

@@ -16,8 +16,7 @@ packages and releases. Do not edit or commit files in that directory. See the
 commands.
 
 Model-based evaluations live in [`evals/`](evals/README.md), outside the plugin
-source. Deterministic triage checks and the MCP reducer IPC regression remain
-part of normal CI.
+source. Deterministic triage checks remain part of normal CI.
 
 Search [existing issues](https://github.com/openai/codex-security/issues)
 before opening a new one.

docs/project-configuration.md

@@ -348,3 +348,11 @@ and applies its severity policy to recovered findings.
 Workflow identity records explicitly requested deep settings, not ambient values
 or shipped defaults, so changing those defaults does not prevent resumption.
 Changing the explicit request still requires a different workflow ID.
+
+### Native executable selection
+
+Native plugin sessions use `CODEX_CLI_PATH` when supplied, or a real `codex`
+executable on `PATH` (`codex.exe` on Windows). The selected executable must be
+outside the scan target. Windows npm shims, managed-package directories and
+desktop cache directories are no longer searched; set `CODEX_CLI_PATH` to the
+installed executable when it is not directly on `PATH`.

evals/README.md

@@ -5,9 +5,5 @@ being embedded in its source tree or shipped npm runtime.
 
 - [Triage finding](triage-finding/README.md): Promptfoo input contracts, OSS
   calibration, and SastBench. This suite owns its private package and lockfile.
-- [Deep reducer](deep-reducer/README.md): optional model evaluation of real IPC
-  size-limit recovery, pagination, and finding retention. It reuses the MCP
-  test helpers and dependencies.
 
-Model runs are opt-in. CI still runs the deterministic triage helper checks and
-the real-IPC reducer regression through the normal MCP test suite.
+Model runs are opt-in. CI still runs the deterministic triage helper checks.

plugins/codex-security/mcp-app/main.ts

@@ -1,25 +1,16 @@
 import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";
-import { createCodexSecurityArtifactWriterServer } from "./artifact-writer-main.js";
 import { createCodexSecurityServer } from "./server.js";
 
 async function main(): Promise<void> {
-  const artifactWriter = process.argv.includes("--artifact-writer");
-  const server = artifactWriter
-    ? await createCodexSecurityArtifactWriterServer()
-    : createCodexSecurityServer();
+  const server = createCodexSecurityServer();
   await server.connect(new StdioServerTransport());
   let closing = false;
   const close = async (exitCode?: number): Promise<void> => {
     if (closing) return;
     closing = true;
     if (exitCode !== undefined) process.exitCode = exitCode;
     await server.close().catch((error: unknown) => {
-      console.error(
-        artifactWriter
-          ? "Codex Security artifact writer failed to close:"
-          : "Codex Security MCP server failed to close:",
-        error,
-      );
+      console.error("Codex Security MCP server failed to close:", error);
     });
   };
   process.stdin.once("end", () => void close());
@@ -28,11 +19,6 @@ async function main(): Promise<void> {
 }
 
 main().catch((error) => {
-  console.error(
-    process.argv.includes("--artifact-writer")
-      ? "Codex Security artifact writer failed to start:"
-      : "Codex Security MCP server failed to start:",
-    error,
-  );
+  console.error("Codex Security MCP server failed to start:", error);
   process.exitCode = 1;
 });

plugins/codex-security/mcp-app/package.json

@@ -11,14 +11,13 @@
   },
   "dependencies": {
     "@modelcontextprotocol/sdk": "^1.30.0",
-    "@openai/codex-sdk": "0.158.0",
-    "smol-toml": "1.8.0",
     "zod": "^4.6.5"
   },
   "devDependencies": {
     "@types/node": "^26.6.2",
     "esbuild": "^0.28.2",
     "prettier": "3.9.8",
+    "smol-toml": "1.8.0",
     "typescript": "^7.0.2"
   }
 }