The Agent Skills Directory

[PROMPT_INJECTION] (SAFE): No override or bypass markers were detected. The skill instructions are purely task-oriented for project navigation.
[DATA_EXFILTRATION] (SAFE): The skill uses tools to read project source and documentation files. No access to system secrets or network capabilities were found.
[INDIRECT_PROMPT_INJECTION] (LOW):
Ingestion points: Reads external memory files (EP{编号}.md) and user-provided EP identifiers.
Boundary markers: None explicitly defined.
Capability inventory: Uses mcp__serena__read_memory, Read, and Grep. These are strictly read-only tools.
Sanitization: None present.
Risk: Although it processes external data, the lack of write or execute capabilities limits the potential for exploitation.
[REMOTE_CODE_EXECUTION] (SAFE): No remote script downloads or dynamic execution patterns were identified.

ep-navigator