Shoofly
Runtime security for AI agents that blocks prompt injection, malware, credential theft, unauthorized writes, and runaway loops before tool calls fire
The product is runtime security software for AI coding agents that blocks prompt injection, malware embedded in tool results, credential theft, unauthorized file writes, and runaway agent loops before tool calls execute. It sells to developers and IT teams in businesses that run Claude Code, OpenClaw, Cowork and Dispatch agents. It is positioned as a local, policy-as-code layer with a free detection and alerting tier and a paid tier that intercepts threats pre-execution, with audit logging to local SQLite and YAML rules.
Key features
- Blocks prompt injection (8 patterns)
- Blocks malware in tool results
- Blocks credential theft
- Blocks unauthorized writes
- Stops runaway agent loops
- Scans for secrets at rest
- Local SQLite audit trail
- YAML policy-as-code
- Multi-channel alerts
- Pre-execution blocking
- 100% local detection
- Custom block/allow policies
- No social media activity within the last 30 days
GTM channels
- Blog
- API
- Docs
ICP
- Software developers
- Security teams
- IT teams