# BLAZLE (blazle.io)

Intercepts AI agent prompts, shell commands and code diffs before they leave, redacting secrets and blocking injections.

- Atlas card: https://inite.ai/en/atlas/blazle.io
- Site: https://blazle.io
- For: Product developers and teams whose AI agents handle prompts, shell commands and code diffs
- Categories: dev
- Use not measured/100 · alive 100/100

## Used for

- redact secrets from prompts before they reach models
- block destructive shell commands issued by coding agents
- scan AI-generated code diffs for committed credentials

## Price

- from 20 USD/month, free plan (read 2026-10-29 on https://blazle.io/pricing)

## For developers

- API docs: https://blazle.io/docs/api-prompt
- Read 2026-10-29

## Your data

- Certifications stated: soc2, iso27001
- Training on your data: no (https://blazle.io/trust)
- Read 2026-10-29

---

Source: INITE Atlas, https://inite.ai/en/atlas/blazle.io. JSON: https://inite.ai/api/atlas/blazle.io. MCP: https://inite.ai/api/mcp (atlas_lookup).