cohort-de-marketing-claude-skills
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installation involves cloning a repository from GitHub (github.com/marketingLendario/cohort-de-marketing.git) to provide the necessary skill definitions and templates.
- [INDIRECT_PROMPT_INJECTION]: Several skills, specifically
/espiao-do-concorrenteand/trend-hunting, are designed to ingest and process data from external sources such as the Meta Ad Library, Google Ads, customer reviews, Reddit, and Twitter. This creates a surface for indirect prompt injection if malicious instructions are embedded in the retrieved content. - Ingestion points: Slash command inputs for competitor and trend research (SKILL.md).
- Boundary markers: No explicit boundary delimiters or instructions to ignore embedded commands are specified in the processing workflow.
- Capability inventory: The skill suite has the ability to write files (Markdown, HTML, PDF, DOCX) and perform network operations via optional API keys for Meta, Google, and OpenAI.
- Sanitization: No sanitization or filtering logic is described for handling external data before it is processed by the agent.
Audit Metadata