okrs
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions explicitly prohibit the inclusion of sensitive data, such as credentials, private file paths, or confidential business information, within the OKR artifacts it generates.
- [SAFE]: The skill metadata references a source repository owned by the author (github.com/jacob-balslev/skill-graph), representing legitimate vendor infrastructure for documentation and skill management.
- [INDIRECT_PROMPT_INJECTION]: The skill defines an ingestion surface for user-supplied strategic context which could be used for indirect prompt injection. However, the risk is mitigated by the highly restricted tool environment (limited to Read and Grep) and the enforcement of a specific, structured output template. Evidence Chain: Ingestion points are the strategic source text inputs described in SKILL.md; boundary markers are provided by the rigid output template; the capability inventory is restricted to basic file-reading tools; no explicit sanitization logic is specified for the input data.
Audit Metadata