This release adds 3 notable features for engineering teams evaluating rollout.
✓ No known CVEs patched in this version
Topics
+14 more
Affected surfaces
Summary
AI summaryNew Web Governance Dashboard with real‑time audit streaming and policy editor is introduced.
Full changelog
What's New
🖥️ Web Governance Dashboard
- Real-time SPA dashboard with 7 pages (overview, audit, policy, anomalies, compliance, regulatory, system)
- WebSocket real-time streaming:
/ws/auditstreams audit entries live to connected clients - Policy editor: in-dashboard YAML editor with validate and save/reload (hot-reload)
- Audit JSON export: one-click filtered export from dashboard audit page
- Auto-refresh overview (30s, toggleable)
aegis serve policy.yaml --seed-demo 200populates demo data for evaluation
🎮 Interactive Playground
- Browser-only policy playground — no install, no backend
- YAML + glob matching in pure JS, try policies instantly
🏢 Enterprise Features
- Cryptographic audit chain: SHA-256/SHA3-256 hash-linked, tamper-evident (EU AI Act Art.12 + SOC2 CC7.2)
- Regulatory compliance mapper: EU AI Act (10 req), NIST AI RMF (8 req), SOC2 (6 req), ISO 42001
- Behavioral anomaly detection: rate spike, burst, new-action, unusual target, auto policy generation
- Compliance report generator: SOC2, GDPR, governance from audit logs
- RBAC: 12 permissions, 5 hierarchical roles, thread-safe AccessController
- Multi-tenant isolation: TenantContext, TenantRegistry, quota enforcement
- Policy versioning: git-like commit, diff, rollback, tagging
- Rate limiter: per-agent and global sliding-window
- Webhook notifications: Slack, PagerDuty, generic JSON
🛡️ CI/CD Governance
- GitHub Action:
aegis-actionfor policy enforcement in PR and deploy pipelines - Shields.io badge:
GET /api/v1/badge/scorefor README embedding
🔍 AI-Specific Deep Features
- Semantic conditions engine (keyword + pluggable LLM evaluator)
- Agent trust chain (hierarchical identity, delegation, cascade revocation)
- Policy diff & impact analysis (
aegis diff) - Natural language autopolicy (
aegis autopolicy) - Adversarial probe (
aegis probe) - Policy testing framework with auto-generation
- Action replay & simulation engine
📊 Stats
- 1,776+ tests passing
- 27 core modules, 65 source files
mypy --strictclean- CI green on Python 3.11, 3.12, 3.13
Full Changelog: https://github.com/Acacian/aegis/compare/v0.1.8...v0.2.0
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About Acacian/aegis
Policy-based governance for AI agent tool calls. YAML policies, approval gates, risk assessment, and audit logging. Cross-platform: LangChain, OpenAI, Anthropic, MCP.
Related context
Beta — feedback welcome: [email protected]