diff --git a/.coderabbit.yaml b/.coderabbit.yaml
new file mode 100644
index 00000000..86367ec3
--- /dev/null
+++ b/.coderabbit.yaml
@@ -0,0 +1,148 @@
+# CodeRabbit AI Configuration for snapm
+#
+# This configuration ensures CodeRabbit focuses on actionable bug fixes
+# and security issues rather than incomplete or theoretical improvements.
+#
+# Schema: https://coderabbit.ai/integrations/schema.v2.json
+# Validate: comment "@coderabbitai configuration" on any PR
+
+language: "en-US"
+tone_instructions: "Security-focused. State risk severity and impact."
+
+reviews:
+  profile: "assertive"
+  request_changes_workflow: true
+  high_level_summary: true
+  collapse_walkthrough: false
+
+  path_filters:
+    - "!vendor/**"
+    - "!node_modules/**"
+    - "!dist/**"
+    - "!build/**"
+    - "!*.min.js"
+    - "!.venv/**"
+
+  auto_review:
+    enabled: true
+    auto_incremental_review: true
+    drafts: false
+    base_branches:
+      - "main"
+
+  # Path-specific rules
+  path_instructions:
+
+    # ── Repo-wide review guidance ────────────────────────────────
+    - path: "**/*"
+      instructions: |
+        - Only suggest refactoring or API changes if the suggestion includes
+          complete implementation (e.g., if suggesting dependency injection,
+          include the test code that uses it)
+        - Avoid suggesting changes that add complexity without immediate,
+          demonstrable value
+        - Focus on bugs, security issues, and correctness problems in PR reviews
+        - Do not suggest changes to code marked with '# pragma: no cover'
+          unless fixing an actual bug
+        - Avoid premature optimization and theoretical improvements
+        - When creating issues for potential improvements, ensure they describe
+          complete solutions, not partial API changes
+
+    # ── Python security ──────────────────────────────────────────
+    - path: "**/*.py"
+      instructions: |
+        Python security:
+        - Prioritize correctness and security over style and architecture
+        - Flag actual bugs, race conditions, and security vulnerabilities
+        - Avoid suggesting testability improvements without corresponding test implementations
+        - SQL: parameterized queries only; no string concatenation
+        - Command: no shell=True, os.system with user input
+        - Path traversal: canonicalize paths, reject ../
+        - Deserialization: no pickle.loads on untrusted data
+        - Integer overflow: bounds-check user-supplied sizes
+        - Never ignore error returns
+
+    # ── IDE & AI tool configuration (supply chain risk) ─────────
+    - path: "**/{.claude,.vscode}/**/*"
+      instructions: |
+        HIGH RISK — IDE and AI tool configuration:
+        These directories can be supply chain attack vectors.
+        Review every change with a security and malware lens.
+        Flag the following for close scrutiny:
+        - .claude: hook commands, MCP server definitions, or
+          permissive tool-use policies in settings.json
+        - .claude: any scripts or binaries (e.g., .mjs, .sh)
+        - .vscode: tasks.json or launch.json entries that execute
+          code on folder open, build, or save
+        - .vscode: settings.json entries that auto-run formatters,
+          linters, or extensions with broad permissions
+        - Any change that grants broad filesystem or network access
+        - Obfuscated content or suspiciously large files
+        - Changes without a clear, legitimate purpose
+
+    # ── CI/CD & GitHub Actions ───────────────────────────────────
+    - path: ".github/workflows/**/*"
+      instructions: |
+        CI/CD security:
+        - Pin actions by full SHA, not tag
+        - No secrets in logs; mask sensitive outputs
+        - Least privilege: minimize GITHUB_TOKEN permissions
+        - No pull_request_target with checkout of PR head
+        - SAST/SCA steps in pipeline
+
+  # ── Security scanners ────────────────────────────────────────
+  tools:
+    gitleaks:
+      enabled: true
+    semgrep:
+      enabled: true
+    actionlint:
+      enabled: true
+
+  # ── Pre-merge checks (hard gates) ───────────────────────────
+  pre_merge_checks:
+    override_requested_reviewers_only: true
+    description:
+      mode: "warning"
+    custom_checks:
+      - name: "no-hardcoded-secrets"
+        instructions: |
+          Flag hardcoded secrets: API keys, tokens, passwords, private
+          keys, credentials. Also flag base64 strings >32 chars in config,
+          URLs with embedded credentials, variables named api_key/secret/
+          token/password assigned string literals.
+        mode: "error"
+
+      - name: "no-weak-crypto"
+        instructions: |
+          Flag MD5, SHA1, DES, RC4, 3DES, Blowfish, ECB mode usage.
+          Flag custom crypto implementations. Flag non-constant-time
+          comparison of secrets or tokens.
+        mode: "error"
+
+      - name: "no-injection-vectors"
+        instructions: |
+          Flag SQL string concatenation, shell=True with user input,
+          eval/exec on untrusted data, pickle.loads on untrusted input,
+          yaml.load without SafeLoader, os.system with variables.
+        mode: "error"
+
+      - name: "no-sensitive-data-in-logs"
+        instructions: |
+          Flag logging that may expose passwords, tokens, API keys,
+          PII (email, SSN, credit card), session IDs, internal
+          hostnames, or customer data.
+        mode: "error"
+
+# ── Knowledge base ───────────────────────────────────────────
+knowledge_base:
+  code_guidelines:
+    enabled: true
+    filePatterns:
+      - "**/CONTRIBUTING.md"
+      - "**/SECURITY.md"
+
+  issues:
+    scope: "auto"
+  pull_requests:
+    scope: "auto"