The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by reading content from local project files that are subsequently used to influence agent behavior.
Ingestion points: The skill uses fast_read_file to ingest content from activeContext.md and progress.md in SKILL.md.
Boundary markers: Absent; there are no explicit instructions to the agent to treat the file content as untrusted data or to ignore embedded instructions.
Capability inventory: The skill possesses file-write capabilities (fast_write_file) and search capabilities (fast_search_files) across the project directory as defined in SKILL.md.
Sanitization: No content validation or sanitization is performed on the data read from the Memory Bank files before it is processed.
[COMMAND_EXECUTION]: The skill performs file system operations using absolute paths within a specific project directory.
Evidence: Operational instructions specify the use of fast_read_file and fast_write_file on paths under /home/kidpixel/render_signal_server-main/.

end-skill