kai-cold-outreach
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill automatically explores the local codebase, reading files such as
package.json,README.md,CLAUDE.md, and other project files to extract marketing and product context. This represents an exposure of project metadata to the agent context. - [COMMAND_EXECUTION]: The instructions reference the use of commands located in a
scripts/directory within the plugin's installation root, indicating the capability to execute local scripts. - [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection by processing untrusted data from the user's project files to generate email content.
- Ingestion points:
MARKETING.md,CLAUDE.md,README.md,PROJECT.md,package.json, and arbitrary project files found during auto-exploration. - Boundary markers: Absent. No specific delimiters or instructions are provided to the agent to ignore potentially malicious content within these files.
- Capability inventory: The skill writes multiple files to the
workspace/outreach/directory and references script execution via thescripts/path. - Sanitization: Absent. There is no evidence of validation or filtering of the content ingested from the codebase before it is used to influence agent output.
Audit Metadata