The Agent Skills Directory

[Data Exposure & Exfiltration] (HIGH): The skill transmits sensitive MEEGLE_USER_KEY and MEEGLE_MCP_KEY as query parameters in the URL within scripts/mcp-proxy.js. Credentials sent via query strings are often leaked in server logs and proxy records.
[Persistence Mechanisms] (HIGH): The scripts/setup.sh script modifies user shell profiles (~/.bashrc, ~/.zshrc, ~/.bash_profile) to store API keys as environment variables. While intended for convenience, this persists sensitive data in plaintext across sessions.
[Indirect Prompt Injection] (HIGH): The skill has a high-severity vulnerability surface.
Ingestion points: Data enters the agent context via API responses from project.larksuite.com processed in scripts/mcp-proxy.js.
Boundary markers: None identified. External content is not delimited from agent instructions.
Capability inventory: The skill can create projects, update tasks, move workflow stages, and manage team members (as described in SKILL.md and README.md).
Sanitization: No sanitization or validation of the external API content is performed before returning it to the agent.
[Privilege Escalation] (MEDIUM): The scripts/setup.sh script automatically applies chmod +x to internal scripts. While common for setup, this modification of file permissions is a privilege-related action.

meegle-mcp