70High
The skill attempts to bypass human-in-the-loop controls and manipulates instruction boundaries to override system prompts, posing a significant risk to agent autonomy and security.
Skill Info
Nameanthropics/knowledge-work-plugins/audit-support
Registrygithub
Version55447a28e6c7
PURLpkg:github/anthropics/knowledge-work-plugins@55447a28e6c7?skill=audit-support
Stars21,418
Installs1,900
Source
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
GitHub
1,900 installs
Skills.shDocs
npx skills add https://github.com/anthropics/knowledge-work-pluginsAssessments (3)
AI Agent Traps ↗Instruction boundary manipulation detectedhighprompt injectionAML.T0051AML.T0051.000AML.T0051.001AML.T0054LLM01:2025
Instruction boundary manipulation detected
100% confidenceline 113
### System
Autonomous operation instructions detected — attempts to remove human-in-the-loop controlsmediumautonomy abuseAML.T0034.002AML.T0053LLM06:2025LLM10:2025
Autonomous operation instructions detected — attempts to remove human-in-the-loop controls
100% confidenceline 324
without human intervention
Missing licenseinfopolicy violation
Skill does not specify a license field. Specifying a license helps users understand usage terms.
100% confidence
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/anthropics/knowledge-work-plugins/audit-support)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/anthropics/knowledge-work-plugins/audit-support"><img src="https://mondoo.com/ai-agent-security/api/badge/github/anthropics/knowledge-work-plugins/audit-support.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/anthropics/knowledge-work-plugins/audit-support.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow