The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing external data for model adaptation.
Ingestion points: Data provided to mcp__claude-flow__ruvllm_microlora_adapt (training data) and mcp__claude-flow__ruvllm_sona_adapt (feedback signals).
Boundary markers: No explicit instructions are provided to the agent to treat this data as untrusted or to use specific delimiters.
Capability inventory: The skill has access to Bash and model configuration tools as listed in SKILL.md.
Sanitization: No validation or sanitization steps for the input data are defined in the instructions.

llm-config