cli-anything-mailchimp
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches the CLI harness from the vendor's GitHub repository (https://github.com/HKUDS/CLI-Anything.git).
- [COMMAND_EXECUTION]: Executes shell commands via the cli-anything-mailchimp tool to interact with the Mailchimp Marketing API.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from the Mailchimp API, which could potentially contain untrusted content that might influence agent behavior. 1. Ingestion points: External data is ingested from the Mailchimp API through commands like reports list, campaigns list-content, and search-members list. 2. Boundary markers: No explicit boundary markers or instructions to ignore embedded content are defined in the skill documentation. 3. Capability inventory: The skill uses the cli-anything-mailchimp tool to execute shell commands for API interaction, including data retrieval and modification. 4. Sanitization: There is no mention of sanitization or validation of the data retrieved from the API before it is presented to the agent.
Audit Metadata