Pipelock Agent Security Scan

Version updated for https://github.com/luckyPipewrench/pipelock to version v1.5.0.

  • This action is used across all versions by 0 repositories.

Action Type

This is a Composite action.

Go to the GitHub Marketplace to find the latest changes.

Action Summary

Pipelock is an open-source firewall designed to protect AI agents by monitoring and blocking unauthorized or malicious API key exfiltration during HTTP requests. It automates the detection of sensitive data leaks, such as API keys, and provides runtime security without requiring code changes, SDKs, or dependencies. Key features include forward proxy and fetch proxy modes, enabling seamless integration with various AI frameworks and HTTP clients.

Release notes

Changelog

  • 3f93984c33143c36a9fab5647b1796b7cfe55053 feat: OTLP log export sink (HTTP/protobuf) (#262)
  • 753a2589880dc14820e93e27dde14481fc187145 feat: adaptive enforcement v2 — escalation-aware enforcement across all transports (#256)
  • 35d831b38c16fb65382fa5a17494860df21bc658 feat: community rules rollout — build wiring, docs, and registry URL (#255)
  • f76467a752af8a66c9a3d87e3d2c8217d0a7a15a feat: filesystem sentinel for subprocess MCP mode (#261)
  • 48bb9397335f64527ee9460132be9edebd8885b3 feat: financial DLP patterns with checksum validation (#258)
  • 66eda7be4f7509ec7e2b640fc5e1b362b9e3627e feat: key-scoped tool policy matching (arg_key) (#257)
  • aca9df9e82a7c4af4d1ed89f62709538e2fec958 fix: adaptive enforcement death spiral (#266)
  • e188cb63293a944c9d06f9c71b158b9c475e4cf4 fix: harden shell normalization against 3 evasion techniques (#259)
  • 3309fddadcab6fa7814884e07834d53682817537 fix: reject unsupported dlp.action and per-pattern action fields (#263) (#264)
  • dda4c33d14255e6300720d005aafae528d1f0d43 fix: transport parity — WS header DLP + forward HTTP response scanning (#254)