https://github.com/Justin0504/aegis

Runtime policy enforcement for AI agents. Cryptographic audit trail, human-in-the-loop approvals, kill switch. Zero code changes.
https://github.com/Justin0504/aegis

ai-agents ai-safety anthropic audit-trail langchain llm-observability mcp policy-engine

Last synced: 10 days ago
JSON representation

Runtime policy enforcement for AI agents. Cryptographic audit trail, human-in-the-loop approvals, kill switch. Zero code changes.

• Host: GitHub
• URL: https://github.com/Justin0504/aegis
• Owner: Justin0504
• License: mit
• Created: 2026-03-04T03:24:33.000Z (5 months ago)
• Default Branch: main
• Last Pushed: 2026-07-15T00:07:08.000Z (11 days ago)
• Last Synced: 2026-07-15T01:22:23.641Z (11 days ago)
• Topics: ai-agents, ai-safety, anthropic, audit-trail, langchain, llm-observability, mcp, policy-engine
• Language: TypeScript
• Homepage:
• Size: 130 MB
• Stars: 365
• Watchers: 22
• Forks: 37
• Open Issues: 4
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • Contributing: CONTRIBUTING.md
  • License: LICENSE
  • Security: SECURITY.md
  • Roadmap: ROADMAP.md

Awesome Lists containing this project

• awesome-auditable-ai - TypeScript - in-the-loop approvals, and a kill switch, applied with no changes to agent code. (Audit Trails and Decision Records / Decision-Record Tools)