sondera-ai/sondera-coding-agent-hooks

README preview

# Coding Agent Hooks by Sondera

> Released as part of the *Hooking Coding Agents with the Cedar Policy Language*
> talk at [Unprompted 2026](https://unpromptedcon.org/), and updated for
> presentation at
> [Black Hat Arsenal 2026](https://blackhat.com/us-26/arsenal/schedule/index.html)
> and the
> [Vegas AI Security Forum '26](https://aisecurity.forum/events/vegas-ai-security-forum-26/).
> See also our ICML 2026 *Agents in the Wild* workshop paper:
> [arXiv:2606.26649](https://arxiv.org/abs/2606.26649).

A reference monitor for AI coding agents. Rust hook binaries and
[Cedar](https://docs.cedarpolicy.com/) policies intercept every shell command,
file operation, and web request to forbid exfiltration and destructive
behaviors, and enforce information flow control. YARA signatures and Cedar
policy evaluation are deterministic. The optional LLM-based classifiers (data
sensitivity, secure code policy) are probabilistic and configurable in
`.sondera/sondera.toml` — see [Configuration](docs/configuration.md).

Works with [Claude Code](https://code.claude.com/docs/en/hooks),

View full repository on GitHub →