continuous-learning
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill operates by extracting patterns from mobile development sessions and coding history to generate new agent skills. This ingestion of untrusted external data (project source code, developer notes) creates an attack surface for indirect prompt injection.
- Ingestion points: Development session data, coding patterns, and architecture decisions (SKILL.md).
- Boundary markers: None identified in the documentation to distinguish between instructions and data.
- Capability inventory: The skill description implies the ability to write to the local file system (
.claude/instincts/) and generate/modify agent skills based on processed data. - Sanitization: No sanitization, filtering, or validation logic for the ingested session data is mentioned.
- [NO_CODE]: No executable scripts (such as Python, JavaScript, or Shell), external dependencies, or dynamic configuration files were provided for analysis. The security posture is evaluated based on the functional descriptions in the markdown documentation.
Audit Metadata