fix(ci): add CodeQL custom sanitizer config to reduce false positives (#1826)

.github/codeql/codeql-config.yml

@@ -0,0 +1,45 @@
+# CodeQL configuration for AutoBot
+# Declares custom sanitizer functions to reduce false positive alerts.
+# See: https://docs.github.com/en/code-security/code-scanning/creating-an-advanced-setup-for-code-scanning/customizing-your-advanced-setup-for-code-scanning
+#
+# Related issue: #1826
+
+name: "AutoBot CodeQL Configuration"
+
+# Model packs that declare custom sanitizers, sources, and sinks
+packs:
+  python:
+    - codeql/python-queries
+
+# Data extensions for custom threat models
+extensions:
+  - addsTo:
+      pack: codeql/python-all
+      extensible: summaryModel
+    data: []
+
+  # ---------------------------------------------------------------
+  # Path-injection sanitizers
+  # ---------------------------------------------------------------
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sinkModel
+    data: []
+
+  # Declare custom validators as sanitizers via the threat-model
+  # extensible. Each entry maps a fully-qualified function to a
+  # sanitizer role for a specific query kind.
+
+# Paths to scan
+paths:
+  - autobot-backend
+  - autobot-shared
+  - autobot-slm-backend
+
+# Paths to ignore
+paths-ignore:
+  - autobot-frontend/node_modules
+  - autobot-backend/venv
+  - "**/__pycache__"
+  - "**/migrations"
+  - "**/tests"

.github/codeql/extensions/autobot-sanitizers.yml

@@ -0,0 +1,101 @@
+# CodeQL data extensions for AutoBot custom sanitizer functions.
+# These declarations tell CodeQL that certain functions act as
+# sanitizers (they validate/sanitize untrusted input), preventing
+# false-positive alerts for code paths that pass through them.
+#
+# Reference: https://codeql.github.com/docs/codeql-cli/using-custom-queries-with-the-codeql-cli/#creating-a-data-extension
+# Related issue: #1826
+
+extensions:
+  # -----------------------------------------------------------------
+  # Path-injection sanitizers
+  # -----------------------------------------------------------------
+  # autobot_shared.security.path_validator.validate_path()
+  # Validates and resolves paths, rejects traversal and symlinks.
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sanitizerModel
+    data:
+      - ["autobot_shared.security.path_validator", "Member[validate_path].ReturnValue", "path-injection"]
+
+  # api/files.py _validate_admin_path()
+  # Restricts paths to allowed admin directories.
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sanitizerModel
+    data:
+      - ["api.files", "Member[_validate_admin_path].ReturnValue", "path-injection"]
+
+  # api/filesystem_mcp.py _validate_file_path() and _validate_directory_path()
+  # Resolve and constrain file/directory paths to allowed roots.
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sanitizerModel
+    data:
+      - ["api.filesystem_mcp", "Member[_validate_file_path].ReturnValue", "path-injection"]
+      - ["api.filesystem_mcp", "Member[_validate_directory_path].ReturnValue", "path-injection"]
+
+  # api/codebase_analytics/endpoints/ownership.py _validate_path_security()
+  # Validates paths against the project root to prevent traversal.
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sanitizerModel
+    data:
+      - ["api.codebase_analytics.endpoints.ownership", "Member[_validate_path_security].ReturnValue", "path-injection"]
+
+  # api/codebase_analytics/endpoints/environment.py _validate_env_path_security()
+  # Validates environment file paths against the project root.
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sanitizerModel
+    data:
+      - ["api.codebase_analytics.endpoints.environment", "Member[_validate_env_path_security].ReturnValue", "path-injection"]
+
+  # api/analytics_evolution.py _validate_evolution_repo_path()
+  # Validates repository paths for evolution analytics.
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sanitizerModel
+    data:
+      - ["api.analytics_evolution", "Member[_validate_evolution_repo_path].ReturnValue", "path-injection"]
+
+  # -----------------------------------------------------------------
+  # SSRF sanitizer
+  # -----------------------------------------------------------------
+  # autobot_shared.security.input_sanitizer.validate_url()
+  # Validates URLs against allowlists, rejects internal/private IPs.
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sanitizerModel
+    data:
+      - ["autobot_shared.security.input_sanitizer", "Member[validate_url].ReturnValue", "request-forgery"]
+
+  # -----------------------------------------------------------------
+  # LDAP-injection sanitizer
+  # -----------------------------------------------------------------
+  # autobot_shared.security.input_sanitizer.sanitize_ldap_filter()
+  # Escapes LDAP special characters in filter values.
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sanitizerModel
+    data:
+      - ["autobot_shared.security.input_sanitizer", "Member[sanitize_ldap_filter].ReturnValue", "ldap-injection"]
+
+  # -----------------------------------------------------------------
+  # Command-injection sanitizers (Pydantic field validators)
+  # -----------------------------------------------------------------
+  # api/git_mcp.py Pydantic @field_validator decorators on request
+  # models validate and sanitize file_path, repo_path, commit, and
+  # ref fields before they reach subprocess calls.
+  - addsTo:
+      pack: codeql/python-all
+      extensible: sanitizerModel
+    data:
+      - ["api.git_mcp", "Member[GitLogRequest].Member[file_path]", "command-injection"]
+      - ["api.git_mcp", "Member[GitLogRequest].Member[repo_path]", "command-injection"]
+      - ["api.git_mcp", "Member[GitDiffRequest].Member[file_path]", "command-injection"]
+      - ["api.git_mcp", "Member[GitDiffRequest].Member[commit]", "command-injection"]
+      - ["api.git_mcp", "Member[GitDiffRequest].Member[repo_path]", "command-injection"]
+      - ["api.git_mcp", "Member[GitBlameRequest].Member[file_path]", "command-injection"]
+      - ["api.git_mcp", "Member[GitBlameRequest].Member[repo_path]", "command-injection"]
+      - ["api.git_mcp", "Member[GitShowRequest].Member[ref]", "command-injection"]

.github/workflows/codeql.yml

@@ -0,0 +1,51 @@
+# CodeQL Analysis workflow
+# Uses custom configuration to declare project-specific sanitizers
+# and reduce false-positive security alerts.
+#
+# Related issue: #1826
+
+name: CodeQL Analysis
+
+on:
+  push:
+    branches: [main, Dev_new_gui]
+  pull_request:
+    branches: [main, Dev_new_gui]
+  schedule:
+    # Run CodeQL analysis weekly on Sundays at 3 AM UTC
+    - cron: "0 3 * * 0"
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+permissions:
+  actions: read
+  contents: read
+  security-events: write
+
+jobs:
+  analyze:
+    name: Analyze Python
+    runs-on: self-hosted
+    timeout-minutes: 30
+
+    strategy:
+      fail-fast: false
+      matrix:
+        language: [python]
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v6
+
+      - name: Initialize CodeQL
+        uses: github/codeql-action/init@v3
+        with:
+          languages: ${{ matrix.language }}
+          config-file: .github/codeql/codeql-config.yml
+
+      - name: Perform CodeQL Analysis
+        uses: github/codeql-action/analyze@v3
+        with:
+          category: "/language:${{ matrix.language }}"