The Agent Skills Directory

[SAFE]: No malicious code, obfuscation, or unauthorized data access patterns were found. The skill instructions are limited to localized text manipulation for stylistic consistency and do not involve suspicious network or system operations.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted documents, creating an ingestion surface for potential instructions. 1. Ingestion points: Input document content is scanned for uppercase strings (2-5 letters) as defined in SKILL.md. 2. Boundary markers: No explicit delimiters are used to isolate untrusted data from instructions. 3. Capability inventory: The skill uses the microsoft_docs_search tool and text replacement capabilities. 4. Sanitization: Input content is not sanitized before processing. Risk Assessment: The highly specific regex-like matching criteria and the benign nature of the actions (acronym expansion) prevent external content from effectively hijacking the agent's intent.

acronym-check