70High
The skill is vulnerable to RAG/memory poisoning
Behavior Analysis
Mismatch detected — The stated purpose includes 'query organization structure', but the skill's explicit shortcuts only cover searching for individual users and retrieving their information, not directly querying the organizational hierarchy.
Claims to do
contact (v1): **CRITICAL — 开始前 MUST 先用 Read 工具读取 [`../lark-shared/SKILL.md`](../lark-shared/SKILL.md)，其中包含认证、权限处理**
Actually does
The skill acts as an interface to the `lark-cli` command-line tool, specifically its `contact` subcommand. It provides shortcuts to execute `lark-cli contact +search-user` for searching employees and `lark-cli contact +get-user` for retrieving user details. It also allows running `lark-cli contact --help`.
Skill Info
Namelarksuite/cli/lark-contact
Registrygithub
Version5fa68ccaa04e
PURLpkg:github/larksuite/cli@5fa68ccaa04e?skill=lark-contact
Stars7,923
Installs56,800
Source
SHA-256807a24a43285...
SHA-1df45f2da2fab...
MD5afcbae6f15cc...
TLSH101128825c97...
Size1,053 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
Skills.shDocs
npx skills add https://github.com/larksuite/cli --skill lark-contactAssessments (2)
AI Agent Traps ↗Mandatory pre-task file read for authentication/permissionshighrag poisoning
The agent is explicitly instructed to read `../lark-shared/SKILL.md` before execution, with the file described as containing 'authentication, permission handling'. This could be used to poison the agent's RAG/memory with attacker-controlled instructions, manipulate its understanding of permissions, or harvest credentials if the file is compromised. The 'CRITICAL' and 'MUST' are semantic manipulation tactics.
100% confidenceline 13
**CRITICAL — 开始前 MUST 先用 Read 工具读取 [`../lark-shared/SKILL.md`](../lark-shared/SKILL.md)，其中包含认证、权限处理**
Description Mismatchlowdescription mismatch
The stated purpose includes 'query organization structure', but the skill's explicit shortcuts only cover searching for individual users and retrieving their information, not directly querying the organizational hierarchy.
85% confidence
Stated purpose: '查询组织架构' (query organization structure). Shortcuts: `+search-user`, `+get-user` (search users, get user info).
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/larksuite/cli/lark-contact)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/larksuite/cli/lark-contact"><img src="https://mondoo.com/ai-agent-security/api/badge/github/larksuite/cli/lark-contact.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/larksuite/cli/lark-contact.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