The Agent Skills Directory

[SAFE]: The skill consists of Markdown files (SKILL.md, buzzword-dictionary.md, classification-criteria.md) that define intent detection, complexity assessment, and routing logic. No malicious code, prompt injection overrides, or persistence mechanisms were found.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input to classify intent and suggest routing. While this is an ingestion surface, the risk is negligible as the skill defines no exploitable capabilities such as tool calls, shell execution, or file writes. * Ingestion points: User requests as described in SKILL.md. * Boundary markers: Absent; there are no delimiters or instructions to ignore embedded commands. * Capability inventory: None; no tools or scripts are defined within the skill. * Sanitization: Absent; no input filtering or validation logic is present.

request-analyzer