ctxora-continuous-learning
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill manages a persistent memory framework that saves data from the repository and user feedback. This creates a potential surface where untrusted data from the workspace could be stored and later used to influence agent decisions via memory retrieval.
- Ingestion points: Untrusted data enters the context through repository evidence, validation results, and user decisions before being passed to
memory_save. - Boundary markers: The instructions lack explicit boundary markers or delimiters to differentiate between stored data and active instructions in the memory context.
- Capability inventory: The skill utilizes several capabilities including
memory_save,memory_search,skill_feedback, andskill_learning_status. - Sanitization: The skill includes explicit instructions for the agent to manually filter out secrets, transient output, and guesses, which mitigates but does not eliminate the risk of processing malicious data.
Audit Metadata