发现优质的 AI Agent 技能

来自 openclaw/skills 技能

Release skills to ClawhHub through the full publication pipeline — auto-scaffolding, OPSEC scan, dual review (agent + user), force-push release, security scan verification.

**Skill under test:** `agent-email-inbox` **Skill type:** Technique + Discipline (security enforcement) **Test approach:** Pressure scenarios for security enforcement + application scenarios for technical implementation. Pressure scenarios designed so the wrong answer is genuinely tempting.

**Date:** 2026-02-24 **Version:** 1.0.0 (pre-publication) **Scan Type:** ClawHub Security Evaluation **Analyst:** Navi (automated review)

**Version:** 1.3.6 **Date:** 2026-03-01 **Assessment Type:** Metadata fix (environment variable declaration) **Previous Version:** 1.3.5 (BENIGN code, SUSPICIOUS scan due to metadata)

**Version:** 1.3.4 **Date:** 2026-03-01 **Assessment Type:** Release (two enhancements) **Previous Version:** 1.3.3 (BENIGN) **Issues:** #35, #36

**Version:** 1.3.3 **Date:** 2026-02-28 **Assessment Type:** Feature enhancement (session-specific archiving) **Previous Version:** 1.3.2 (BENIGN) **Issue:** #34

**Version:** 1.3.2 **Date:** 2026-02-28 **Assessment Type:** UX enhancement (human-readable token sizing) **Previous Version:** 1.3.1 (BENIGN) **Issue:** #33

**Version:** 1.3.0 **Date:** 2026-02-28 **Assessment Type:** Feature addition (dynamic context limit detection) **Previous Version:** 1.2.1 (BENIGN) **Issue:** #32

**Assessment Date:** 2026-02-28 (AFTER implementation, BEFORE publication) **Assessed Against:** ClawHub Security Evaluator Criteria **Issue:** #31 - Archive command crashes with multi-agent sessions

**Assessment Date:** 2026-02-28 (BEFORE publication) **Assessed Against:** ClawHub Security Evaluator Criteria **Issue:** #30 - Multi-agent session discovery and unified monitoring

**Assessment Date:** 2026-02-28 (BEFORE publication) **Assessed Against:** ClawHub Security Evaluator Criteria **Issue:** #29 - Make timeout and refresh intervals user-configurable