battlecard
Pass
Audited by Gen Agent Trust Hub on Jul 28, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes content from external, user-provided URLs via
swan-fetch-scraped-url, creating a surface for indirect prompt injection where malicious instructions on a scraped website could attempt to influence the agent. - Ingestion points: Competitor homepages, pricing pages, and case studies are scraped in Step 3 of
SKILL.md. - Boundary markers: No specific delimiters or "ignore instructions" warnings are present to isolate the scraped content from the system prompts.
- Capability inventory: The skill utilizes
swan-update-knowledge-competitionto persist the processed data into the organizational memory. - Sanitization: There is no evidence of filtering or validation of the scraped text before it is used to generate the battlecard content.
- [SAFE]: All tools used (e.g.,
swan-get-memory,swan-fetch-scraped-url,swan-update-knowledge-competition) are standard resources provided by the vendor 'swan-gtm' and are used for their intended research and data management purposes.
Audit Metadata