marketing-skills-guide
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (MEDIUM): The skill identifies and directs users/agents to external repositories and URLs not belonging to trusted organizations. Specifically, it references
sickn33/antigravity-awesome-skills,coreyhaines31/marketingskills, andgmh5225/awesome-skills. These sources are unverifiable and may host malicious or unauthorized skills. - [PROMPT_INJECTION] (LOW): Indirect Prompt Injection surface. The skill explicitly instructs the agent to use
WebFetchto retrieve a full resource list from a raw GitHub URL (gmh5225/awesome-skills). This untrusted external data is ingested into the agent's context without sanitization or boundary markers. - Ingestion points:
SKILL.md(Full Resource List section). - Boundary markers: Absent; the agent is simply told to retrieve and use the file.
- Capability inventory: The agent uses WebFetch to read external content and is then expected to provide the "latest information" based on it.
- Sanitization: Absent; no validation of the fetched Markdown content is performed before processing.
Audit Metadata