sales-data-hygiene
Pass
Audited by Gen Agent Trust Hub on Apr 29, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to use
WebFetchto downloadSKILL.mdfiles from the vendor's official GitHub repository (github.com/sales-skills/sales). This is a legitimate functional requirement to gather context for related skills (e.g.,/sales-mailshake,/sales-klaviyo) before recommending them to the user. Since these downloads target the author's own verified infrastructure for documentation purposes, the activity is considered safe. - [COMMAND_EXECUTION]: The documentation mentions the command
npx skills add sales-skills/sales --skill sales-do. This is the standard installation procedure for the vendor's router skill and does not represent an unauthorized or malicious command execution. - [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by ingesting external content through
WebFetchto inform its logic. - Ingestion points: External
SKILL.mdfiles retrieved fromraw.githubusercontent.com/sales-skills/sales/(as described inSKILL.md). - Boundary markers: Absent; the skill does not specify the use of delimiters or 'ignore' instructions for the fetched content.
- Capability inventory: The skill uses
WebFetchandReadto gather context and recommends specific CLI installation commands and platform configurations. - Sanitization: Absent; the content is fetched and read directly to inform the agent's recommendation strategy.
Audit Metadata