content-amplifier
Pass
Audited by Gen Agent Trust Hub on Jul 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a structured workflow for marketing content management. It relies on user-provided transcripts and performance metrics without requesting unnecessary system permissions.
- [REMOTE_CODE_EXECUTION]: The skill includes explicit security-positive instructions in
references/atom-extraction.md, directing the agent not to install or call external libraries (e.g., whisper, mediapipe, pandas) and to focus solely on text processing. - [DATA_EXPOSURE]: Data handling is restricted to standard agent memory paths (
memory/influencer/content-amplifier/). The skill prompts for user confirmation before promoting facts to durable memory tiers, adhering to best practices for data persistence. - [INDIRECT_PROMPT_INJECTION]: While the skill is designed to ingest and process untrusted text (transcripts and captions), it lacks the capabilities (such as arbitrary network operations or system command execution) required to weaponize an indirect injection. The risk is considered minimal and inherent to the skill's purpose.
Audit Metadata