This release includes 2 security fixes for security teams reviewing exposed deployments.
Published 2mo
MCP Security & Auth
✓ No known CVEs patched
This release patches 2 known CVEs
Topics
agent-security
ai-agent-security
ai-agents
ai-governance
ai-safety
ai-security
+14 more
audit-trail
compliance
guardrails
langchain
llm-security
mcp
mcp-security
model-context-protocol
pii-detection
policy-as-code
policy-engine
policy-testing
prompt-injection
selection-governance
Affected surfaces
auth
rbac
Summary
AI summaryFixed privilege escalation and resource consumption gaming vulnerabilities.
Full changelog
Quality hardening for Selection Governance
Fixed
- Token-boundary keyword matching — prevents false positives (e.g. "undelete" ≠ "delete")
- CongruenceChecker priority — deterministic DELETE > WRITE > READ
- Privilege escalation gaming — no longer blindly concatenates param values
- Resource consumption gaming —
max(count, limit, batch_size) - PII keyword list — synced with design doc (added email, name)
Added
- SelectionAuditor thread safety (
threading.Lock) - CommitRevealSelection TTL + max_pending (memory leak prevention)
@audit_selectionsync function support- 33 new tests
Changed
- GitHub Action renamed: "Aegis AI Agent Security Gate"
See CHANGELOG.md for full details.
Security Fixes
- Prevented privilege escalation by disallowing blind concatenation of parameter values
- Limited resource consumption with `max(count, limit, batch_size)` logic
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]