This release includes 5 security fixes for security teams reviewing exposed deployments.
Published 2mo
MCP Security & Auth
✓ No known CVEs patched
This release patches 5 known CVEs
Topics
agent-security
ai-agent
ai-firewall
ai-safety
ai-security
claude-code
+14 more
cursor
data-exfiltration
dlp
guardrails
hermes-agent
langchain
llm-security
mcp
mcp-security
openclaw
pii-detection
prompt-injection
security
shellward
Affected surfaces
rce_ssrf
auth
Summary
AI summaryReDoS fix speeds email regex from 14s to 43ms on large input.
Full changelog
What's New
Hook Error Handling (Fail-Safe)
createSafeApi()wrapper: all 8 defense layers get automatic try-catchbefore_tool_callerrors → block (deny on error, safer than allow)- Other hook errors → pass-through (don't break the chain)
- All errors logged at CRITICAL level
Non-Blocking Update Notifications
- Checks npm registry for new versions (24h interval)
- Notification dedup: same version only notified once — won't repeat after user has seen it
- Silent on network failure, cached to avoid repeated timeouts
Remote Vulnerability Database
- 17 real CVEs/GHSAs from NVD and GitHub Security Advisories
- 1 supply chain alert (SANDWORM_MODE campaign)
- 24h cache, graceful fallback to local built-in DB
/check-updatescommand shows vuln DB source and details
Security Fixes
- ReDoS fix: email regex 333x speedup on large text (14s → 43ms on 200KB)
- Injection rule gaps: expanded Chinese rules, added cross-language injection detection (26 total rules)
- Fork bomb regex: fixed pattern broken by command splitting
- Defensive input handling: non-string
toolName/paramsno longer crash - Regex state pollution: fixed global flag leak in
/scan-plugins
Testing
- 100 tests across 3 suites (37 integration + 42 edge cases + 21 update check)
- Performance: 125,000 tool checks/sec, 200KB PII scan in 55ms
Install: openclaw plugins install shellward
Security Fixes
- ReDoS fix: email regex performance improved 333× (14 s → 43 ms on 200 KB input)
- Injection rule gaps expanded with Chinese rules and cross‑language detection (26 total rules)
- Fork bomb regex pattern repaired after command splitting breakage
- Defensive input handling now prevents crashes from non‑string `toolName`/`params`
- Regex state pollution fixed: global flag leak in `/scan-plugins` resolved
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 jnMetaCode/shellward
AI Agent Security Middleware & MCP Server with 8-layer defense including prompt injection detection, DLP data flow tracking, command blocking, and PII detection. 7 MCP tools, zero dependencies.
Related context
Beta — feedback welcome: [email protected]