acceptance-criteria-doc

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses a documented attack surface where it ingests untrusted data and utilizes significant write capabilities.
  • Ingestion points: The skill fetches external content via getJiraIssue (epic description), searchJiraIssuesUsingJql (child issues), getConfluencePage (structural reference docs), and gh (GitHub PR and code research).
  • Boundary markers: There are no explicit instructions to use delimiters or "ignore embedded instructions" warnings when processing these data sources.
  • Capability inventory: The skill utilizes createConfluencePage, editJiraIssue (to append links to issue descriptions), and addCommentToJiraIssue across the epic and all its children.
  • Sanitization: The skill does not define any sanitization, validation, or escaping protocols for the content fetched from Jira or GitHub before it is included in new documentation or issue updates.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 07:59 PM
Security Audit — agent-trust-hub — acceptance-criteria-doc