Config change
Auth
Usage reporting, LangChain optional, IORails enhancements
Upgrade now
Dependencies
Cryptography core dependency
No immediate action
MIT license change
Config change
Auth
Auth for remote HTTP
↳
v0.82.1
(1mo)
—
Next 16 export fix
↳
v0.81.1
(1mo)
—
Tenant-scoped gateway routing
No immediate action
PR comment posting + README restructuring
Review required
Auth
Dependencies
Railway bearer token required
Breaking changes
- License migration from MIT to Apache 2.0; NOTICE file added with attribution requirements
Security fixes
- Completed CodeQL path‑injection remediation (Round 2), closing all related security alerts
- Fixed path injection risk in `readAgentConfigContent` with input validation and boundary checks
Notable features
- Added port 18789 to default AI infrastructure scan ports
- Added AIG vulnerability rules (2026‑03‑20 batch)
- Added 281 new CVE/GHSA entries for OpenClaw components
No immediate action
CLI refactor + PCI DSS compliance
↳
v0.71.1
(2mo)
—
ToolHive removal + OS scan
Review required
Auth
RCE / SSRF
Platform‑agnostic core & DLP model
No immediate action
AI scanning + DeepSeek + outputs
↳
v0.70.6
(2mo)
—
Severity default mapping
Breaking changes
- Removed stream_usage from streaming metadata capture
Notable features
- IORails Input/Output rail engine with parallel execution
- check_async method for standalone rail validation
- OpenAI-compatible server with v1/models endpoint
⚠ Upgrade required
- Removed `idSuffix` field from provider configurations; adjust existing configs accordingly
- Dockerfile and docker.sh updated to support Agent-Scan framework and optimize shallow clone with Docker Compose v2 usage
Notable features
- EdgeOne ClawScan: dedicated, AI-powered security portal for OpenClaw with on-demand health checks and comprehensive scans
- Agent-Scan Framework: multi-agent architecture with specialized sub-agents covering OWASP ASI compliance, SSRF detection, config scanning, vulnerability detection, data leakage, and tool abuse
No immediate action
Native OCI layer parser
↳
v0.64.0
(2mo)
—
Zero‑trust auth model
↳
v0.60.1
(2mo)
—
_meta.tools removal
Breaking changes
- Removed streaming field from configuration
Notable features
- Nemotron reasoning-capable content safety models
- GLiNER for open-source PII detection
- Multilingual refusal messages in content safety rails
Notable features
- Added SYS_ADMIN capability for Chrome sandbox and database indexes
- Updated AI tool protocol scan report with model & scan duration
↳
v3.5-preview
(6mo)
—
MCP-Scan framework + attack methods
Notable features
- LangChain 1.x compatibility
- Content blocks API support for reasoning traces and tool calls
↳
v0.18.0
(6mo)
—
Reasoning trace handling
↳
v0.15.0
(9mo)
—
OpenTelemetry API adoption
↳
v0.14.0
(11mo)
—
Reasoning trace handling
