content-research-writer
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a structured workflow for content creation, including outlining, researching, and drafting. All suggested file operations are limited to a user-specified writing directory (e.g.,
~/writing/). - [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data during the research phase and when processing user drafts.
- Ingestion points: External research sources found via search and user-provided draft files (e.g.,
article-draft.md). - Boundary markers: Absent; the instructions do not explicitly require the agent to use delimiters or ignore instructions within external data.
- Capability inventory: The agent has capabilities to create directories (
mkdir), write files (touch, writing content to.mdfiles), and likely perform web searches. - Sanitization: Absent; there are no specific instructions for filtering or escaping content from external sources. However, because this behavior is central to the skill's legitimate purpose (research and writing) and does not involve high-risk capabilities like shell command execution on that data, the severity is considered safe.
Audit Metadata