Cerbere-AG: A control layer for monitoring and gating AI agent tool calls
freakingmus · reddit · 2026-10-09
An indie developer is building Cerbere-AG, a control layer that sits between AI agents and the tools they invoke, addressing the risk that agents can produce convincing explanations while making dangerous tool calls.
Key capabilities:
- Observe agent activity and tool calls
- Detect potentially risky actions
- Apply policies to allow, flag, or block actions
- Maintain an audit trail for debugging and security reviews
- Insert human approval for high-risk operations
It's free during development (local version stays free; hosted plans will come later). The author is recruiting SDK integrators and 3–5 design partners, especially teams whose agents touch sensitive data or external systems. Open source on GitHub.
More from coding & agent
- OpenAI's math results released as open source RL environments — SergioPaniego · 2026-10-09
- Dev builds two playable games in a week with Claude Haiku and GPT small models, with cost breakdowns — chongdashu · 2026-10-09
- Stop vibe coding to 80%: pool your tokens so one person ships it to 100% — arjunrajlab · 2026-10-09
- dhh prefers Codex as main coding model with Claude secondary, praises Sol 6.1 — npew · 2026-10-09
- Leaked Claude Code Setup: Sonnet Builds, Haiku Swarms, Opus Architects — Arindam_1729 · 2026-10-09
- Users claim Codex harness, not the models, is the bottleneck: zcode 5x faster on same tasks — bdsqlsz · 2026-10-09