Policies
Agentic Code Security
Agentic Code Security: what it does, its label on each agent, what it misses, and examples.
Catalog policy agentic-code-security, version 0.4.1, labelled blocks on Claude Code.
What it does
Trigger: writing agent code or agent config -- Python or TypeScript using AutoGen, CrewAI, LangChain, LangGraph, mem0, the OpenAI Agents or Claude Agent SDK, an MCP server or client (.mcp.json, .cursor/mcp.json, .vscode/mcp.json, claude_desktop_config.json, .codex/config.toml, .gemini/settings.json), docker-compose files for agents. avoid: code execution on the host, unpinned MCP servers and models, shell-reaching tools, approvals switched off, whole-environment and credential-store leaks, TLS verification off, unbounded loops, SQL and eval built from strings, stripped provenance markers. 29 rules in 10 packs -- exec, supply, tools, approval, identity, comms, bounds, prompt-memory, code, provenance -- each pack or rule allow|deny in .chock/agentic-security.json; bounds, prompt-memory and one supply rule start as allow.
On each agent
| Agent | Label | What it does there |
|---|---|---|
| Claude Code | Blocks | Blocks on an agent's file writes and at turn end |
| Cursor | Blocks | Blocks on an agent's file writes and at turn end |
| Codex | Blocks | Blocks on an agent's file writes and at turn end. If the hook errors, the action goes through (not witnessed; the vendor's row is only partly live-run) |
| GitHub Copilot | Blocks | Blocks at turn end only. It does not check a file write when it happens. If the hook errors, the action goes through (not witnessed; the vendor's row is only partly live-run) |
| Devin | Blocks | Blocks at turn end only. It does not check a file write when it happens. If the hook errors, the action goes through (vendor docs only, not witnessed) |
| Grok | Not graded | Not graded yet for Grok |
| goose | Not graded | Not graded yet for goose |
| Antigravity | Not graded | Not graded yet for Antigravity |
| OpenHands | Not graded | Not graded yet for OpenHands |
| Kimi Code | Not graded | Not graded yet for Kimi Code |
| Factory Droid | Not graded | Not graded yet for Factory Droid |
What it misses
This policy does not state what it misses.
Examples from its eval suite
Refused: A file agents/proxy.py runs AutoGen code with use_docker False, judged at commit.
Allowed: A file agents/proxy.py runs code in Docker, or not at all, judged at commit.
More
Built from engine 5a4b6287a46e and catalog 1e60d211c1bb.