The Agent Skills Directory

[SAFE]: The skill architecture utilizes a structured multi-agent approach with defined input/output contracts, which maintains clear boundaries and prevents instruction leakage between specialized roles.
[SAFE]: The utility scripts (manifest-sync.ts, append-loop-result.ts, etc.) are used for maintaining the forsvn environment's metadata and structure. These scripts implement best practices, such as symlink detection and path validation, to ensure operations remain within intended directories.
[SAFE]: The skill does not contain any obfuscated code, hidden instructions, or patterns indicative of prompt injection attacks.
[SAFE]: Network operations via WebSearch and WebFetch are explicitly constrained to research tasks and validating business hypotheses, with no signs of data exfiltration.
[SAFE]: Permission requirements, including the Bash tool, are consistent with the skill's purpose of managing a complex file-based strategic workspace and executing its own maintenance scripts.

prioritize