The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's core logic involves ingesting untrusted user prompts to decide whether to build a deterministic workflow or a more autonomous AI agent.
Ingestion points: User requests and scenario prompts are defined as primary inputs in SKILL.md.
Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the user input it processes.
Capability inventory: The skill acts as a router to high-capability tools such as agent-builder and workflow-builder, and performs discovery via list-agent-capabilities and workflows(action="list").
Sanitization: There is no indication that the user-provided input is sanitized or validated before being used to drive the classification decision.

intent-recognition