kai-partnership
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts to perform quality checks on outreach copy and marketing assets. This behavior is confined to the project's local directory. Evidence: Runs
python scripts/quality_gates/banned_word_check.py <file>andpython scripts/quality_gates/four_us_score.py <file>during Phase 4. - [INDIRECT_PROMPT_INJECTION]: The skill processes content from various project files to automate product discovery, which could be a vector for indirect instructions if those files contain untrusted data.
- Ingestion points: Reads
MARKETING.md,CLAUDE.md,README.md,PROJECT.md, andpackage.jsonfrom the project root. - Boundary markers: No explicit boundary markers or instructions to ignore embedded commands were found for the ingested files.
- Capability inventory: The skill has the capability to execute subprocesses (local quality gate scripts) and write to the filesystem (creating
MARKETING.md). - Sanitization: No explicit sanitization or validation of the ingested file content is mentioned.
Audit Metadata