content-retro
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill sources a local library script (
../lib/preamble.sh) to initialize its execution environment. - [PROMPT_INJECTION]: The skill implements an automated feedback loop where historical performance data from
~/.kai-marketing/content-log.jsonlis used to generate and update instructions in~/.kai-marketing/marketing-defaults.md. This represents a surface for indirect prompt injection if the log source is compromised, though it is the skill's primary intended function. - Ingestion points: Data is ingested from
~/.kai-marketing/content-log.jsonl(SKILL.md). - Boundary markers: No specific delimiters are used when writing the analyzed patterns into the defaults file.
- Capability inventory: The skill has read and write access to files within the user's
~/.kai-marketing/directory. - Sanitization: No explicit sanitization or validation of the log content is described prior to its use in generating new instructions.
Audit Metadata