Jev in production › Agent guardrails and approvals
CallSitter
Jev classifies live call transcripts as scam, spam, or genuine with score thresholds of 0.70-0.85 to steer an AI phone agent in real time.
The source publishes no measured number.
- Use
- Agent guardrails and approvals
- Industry
- Security
- Form
- Open-source tool
- Stage
- Beta
- Listed
- 2026-10-19
- Found via
- github
Badge
For the project's own README, linking back here:
Also used for agent guardrails and approvals
- Jev Guard (leepokai) $0.00004
- Canny 25 of 25
- jev-kit 0.0008 USD
- jev-permission-gate 2x faster decisions and zero unsafe allows across 93 labeled test cases
- dsh-jev-interceptor $0.00002 per request
- toolgate 77%
- dsh-jev 25/25 fixture agreement
- jev-engineering (eugeniughelbur) 13 of 13