The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute local utility scripts such as skills/forsvn-slop/scan.ts and scripts/log-critic-override.ts to perform its core auditing functions.
[PROMPT_INJECTION]: The skill processes untrusted marketing draft data, representing an indirect prompt injection surface.
Ingestion points: Untrusted content enters through marketing artifacts located in docs/forsvn/artifacts/**/*.md.
Boundary markers: The skill uses a structured report template and a protected_tokens mechanism to separate user content from instructions.
Capability inventory: The skill possesses the Bash, Skill, and Read tools, enabling it to execute local scripts and dispatch sub-agents.
Sanitization: Security is maintained through a 're-verify gate' that re-scans fixed content to ensure that critical tokens (names, URLs, numbers) are preserved and that no new slop patterns were introduced.
[SAFE]: The auxiliary script log-critic-override.ts demonstrates excellent security hygiene.
It implements directory traversal prevention by rejecting artifact paths containing .. or absolute paths.
It protects against symbolic link attacks by using lstatSync to verify directories and target files before writing.
It performs strict input validation on command-line arguments, including regex checks for skill names and newline detection to prevent injection into markdown logs.

audit-marketing