hermes-self-audit定期审计 Hermes Agent 的自改进行为 - skill 变更、curator 状态、记忆库健康度。防止 Agent 在用户不知情的情况下悄悄进化。建议配合 cronjob 每周运行一次。
Install via ClawdBot CLI:
clawdbot install freepengyang/hermes-self-auditGrade Fair — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Calls external URL not in known-safe list
https://clawhub.ai/freepengyang/hermes-self-auditAudited May 10, 2026 · audit v1.0
Generated Oct 5, 2026
An SRE engineer runs this skill on a weekly cron to audit all skill changes made by the Hermes Agent, including newly created, modified, and archived skills. The report is delivered to Feishu so the team can review hidden autonomous learning before it affects production workflows.
Before deploying a new version of an internal AI agent, release managers use this audit to verify that no unexpected skills were auto-created or that memory provider settings remain unchanged. The readable report provides evidence for change management and rollback decisions.
A support team relies on honcho for long-term customer memory. The audit alerts them if the memory provider gets unintentionally switched to mem0, which could lose user profiles and conversation history continuity. This helps maintain consistent customer experience.
A platform engineering team manages dozens of Hermes agents across teams. They aggregate the weekly audit logs to identify stale skills (e.g., unused for 90+ days) and propose archiving or removal, keeping the shared skill catalog clean and relevant.
In a financial services firm, the audit log serves as a compliance artifact showing that agent self-modifications are monitored and that configuration changes are tracked. The weekly report is stored locally at ~/.hermes/logs/audit.log and pushed to Discord for auditors.
A SaaS platform that installs the Hermes Self-Audit skill on client agents and provides a dashboard aggregating audit reports across all agents. The service sends alerts for anomalies such as unexpected skill creation or curator misbehavior.
Consultants offer workshops and setup services for companies adopting autonomous agents. They configure Hermes Self-Audit, integrate it with existing observability stacks, and train teams to interpret the audit reports for safety and compliance.
A specialized tool that extends the audit skill by providing one-click rollbacks of problematic skill changes and policy enforcement (e.g., auto-approval workflows). It integrates with the audit log to trigger automatic remediation based on detected anomalies.
💬 Integration Tip
Install via `clawhub install hermes-self-audit` and create a weekly cronjob with `hermes cron create` to push reports to your preferred chat platform. Ensure curator is enabled in `~/.hermes/config.yaml` for full audit coverage.
Scored Oct 5, 2026
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope,...
Security scanner for AI agent skills. 9 built-in detection signatures. Identifies secrets, unsafe execution patterns, and prompt injection. Sub-50ms results.
Wallet anti-theft guard. One-click scan for high-risk wallet approvals to protect user assets. Use when a user asks for a wallet security check, wallet healt...
Comprehensive security audit for an agent's full skill stack. Chains scanner, differ, trust-verifier, and health-monitor into a single assessment with priori...
GEO Audit — AI Search Visibility Checker for ChatGPT, Perplexity, Claude & Gemini. 29-point GEO readiness checklist: robots.txt AI crawler access, Index...
Senior SecOps engineer skill for application security, vulnerability management, compliance verification, and secure development practices. Runs SAST/DAST sc...