The Agent Skills Directory

[SAFE]: The skill follows a legitimate auditing workflow, using tools like Grep, Glob, and Bash to analyze codebase structure. It adheres to documented worker contracts for scoring and reporting. No malicious patterns such as credential theft, persistence, or network exfiltration were detected.
[PROMPT_INJECTION]: The skill has an inherent surface for indirect prompt injection because it processes untrusted data from the codebase being scanned.
Ingestion points: Content is read from the scan_path or codebase_root using tools like Read and Grep.
Boundary markers: No explicit delimiters or instructions are used to separate scanned code from the agent's internal logic.
Capability inventory: The skill can read files, execute grep/bash for searching, and write reports to a local output directory. It has no network access.
Sanitization: No sanitization of the ingested codebase content is specified in the workflow.

ln-645-architecture-modernization-auditor