AgentBrake blocks prompt injection exfiltration with verifiable crypto proofs
BOSS_METALLIQUE · reddit · 2026-08-17
AgentBrake is a circuit breaker SDK designed to stop prompt injection data exfiltration in real-time, rather than just detecting it after the fact.
- Core Mechanism: Uses taint tracking to monitor content read by the agent. If the agent attempts to hit an egress tool after reading untrusted content, the call is killed before execution.
- Verifiability: Mints an Ed25519 signed receipt on every trigger, hash-chained to the previous one. Client CISOs can export the bundle and verify offline using only the public key to confirm the block, eliminating the need to trust the server.
- Compliance: Mapped to OWASP Agentic Top 10, designed for the EU AI Act's "demonstrable runtime controls" requirement, generating compliance reports.
- Stack: No proxy, no cloud dashboard, wraps tool calls via a decorator. MIT licensed.
More from coding & agent
- Agents turn conservative as context grows: edits shrink from 150 to 30 lines — dbreunig · 2026-10-16
- ai& becomes an official built-in Cline provider with 13 open-weight models — DavidBennett__ · 2026-10-16
- Open-source rizin-mcp lets LLMs run local binary and malware analysis via Rizin — Pure_Emergency1424 · 2026-10-16
- x402 as a free GTM channel: agents discover 100+ tools, no API keys needed — 0xJeff · 2026-10-16
- Sam Altman: stop getting better at prompts, build loops and graphs that prompt for you — goyalshaliniuk · 2026-10-16
- POV ski game on real Matterhorn elevation data, built with Claude Opus 5.5 — techartist_ · 2026-10-16