client-package-notion

Pass

Audited by Gen Agent Trust Hub on Jul 20, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXFILTRATION]: The skill's primary function involves reading local data from the clients/ directory and transmitting it to external services (Notion and Google Sheets). This behavior is explicitly defined and necessary for the skill's stated purpose of creating shareable client packages.
  • [COMMAND_EXECUTION]: Executes commands via the notion and rube (Composio) MCP servers. These are used to manage workspace pages and perform batch updates to spreadsheets.
  • [PROMPT_INJECTION]: There is an inherent risk of indirect prompt injection as the skill processes content from local Markdown and CSV files.
  • Ingestion points: Contents are read from clients/<client_name>/strategies/, campaigns/, content/, and leads/ files.
  • Boundary markers: The procedure does not specify delimiters or system instructions to ignore potential commands embedded within the client files.
  • Capability inventory: The agent has the ability to write to Google Sheets and create/modify Notion pages across the workspace.
  • Sanitization: While the skill includes steps for escaping special characters for Notion Markdown compatibility, it does not implement security-focused sanitization to filter out malicious instructions in the source text.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 20, 2026, 06:25 PM
Security Audit — agent-trust-hub — client-package-notion