code-refactoring-context-restore
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and 'rehydrate' context from external sources, creating a potential vector for indirect instructions.
- Ingestion points: The skill retrieves data from a
context_source(such as a vector database or file system) and refers to external documentation inresources/implementation-playbook.md(SKILL.md). - Boundary markers: The instructions do not specify any delimiters or safety warnings to prevent the agent from executing instructions found within the restored context.
- Capability inventory: The skill provides logic and commands (e.g.,
context-restore) for retrieving, ranking, and integrating external data into the active agent session (SKILL.md). - Sanitization: There are no explicit sanitization, filtering, or validation steps mentioned to ensure the retrieved context does not contain malicious prompt fragments.
Audit Metadata