The Agent Skills Directory

[SAFE]: No malicious code, obfuscated content, or persistence mechanisms were detected. The skill is composed of natural language instructions and markdown templates.- [SAFE]: The skill incorporates safety-oriented anti-patterns, instructing the user not to publish confidential client details and to verify all claims before inclusion.- [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes external inputs like assignment briefs and project evidence. However, given the purely instructional nature and lack of sensitive system capabilities, this is considered a functional requirement rather than a security flaw. 1. Ingestion points: External assignment briefs, evaluation criteria, and project evidence documents. 2. Boundary markers: The skill does not define specific delimiters for untrusted data. 3. Capability inventory: The skill is scoped to read and search operations. 4. Sanitization: No content sanitization or validation is specified.

05-relevant-experience