rhdh-konflux-tasks
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and fetches content from external sources including the konflux-ci organization's GitHub repositories and Quay.io. These resources provide migration guides and task bundle metadata necessary for the skill's operation. These sources are considered well-known and project-specific.
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted data from external sources (MIGRATION.md files and OCI image layers) to determine code modifications. This is inherent to the skill's maintenance purpose.
- Ingestion points: Fetches migration documentation from GitHub and OCI image metadata (data-acceptable-bundles) from Quay.io.
- Boundary markers: The skill lacks explicit boundary markers in the prompt instructions for the ingested external content, though it follows structured migration logic.
- Capability inventory: The skill performs file system modifications (via sed and yq), network operations (via skopeo), and executes local repository scripts.
- Sanitization: The provided bash scripts use restrictive regular expressions for reference extraction and utilize structured tools like jq and yq for data processing.
Audit Metadata