basecamp-activity
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches activity logs (messages, titles, etc.) from Basecamp, which are external and potentially untrusted sources. This data could contain malicious instructions meant to influence the agent in downstream processing steps.
- Ingestion points: Processed activity events from the
basecamp timelinecommand inscripts/basecamp-activity.sh. - Capability inventory: The script writes to the local file system using
mkdirandjqredirection, and executes external CLI tools. - Boundary markers: None found; data is processed and stored in plain JSON format.
- Sanitization: No content filtering or sanitization is performed on the activity data before caching.
- [COMMAND_EXECUTION]: The script executes multiple system and vendor-specific commands to manage activity data.
- Evidence: Invocation of
basecamp timeline,jq,date,mkdir, andrmwithin the shell script. - [DYNAMIC_EXECUTION]: The script uses dynamic environment setup, which executes output from an external tool manager at runtime.
- Evidence: Use of
eval "$(mise hook-env 2>/dev/null)"inscripts/basecamp-activity.shto initialize the environment for the Basecamp CLI.
Audit Metadata