Stroq has introduced a zero-telemetry "causal firewall" that traces terminal executions by autonomous coding agents back to the untrusted text that provoked them, aiming to block indirect prompt injections before rogue tool calls run. The defense is a TypeScript engine built by Agybay Gibatollayev that intercepts tool calls when agents have shell access, according to the architectural breakdown. The approach matters because unfettered shell access lets routine repository inspections become covert exfiltration channels.
When autonomous coding agents gain unfettered shell access, indirect prompt injections turn routine repository inspections into covert exfiltration channels. Stroq introduces an elegant, zero-telemetry defense by tracing terminal executions directly back to the untrusted text that provoked them. Here is an architectural dissection of how Agybay Gibatollayev’s TypeScript engine intercepts rogue tool calls before they strike.
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.