The Agent Skills Directory

[Prompt Injection] (LOW): The skill demonstrates a vulnerability surface for indirect prompt injection by design, as it instructs the agent to read and process external logs. * Ingestion points: addon.output tool referenced in SKILL.md. * Boundary markers: Absent; there are no instructions to the agent to treat output as untrusted or to use specific delimiters. * Capability inventory: Accesses addon.lint, addon.test, and env.status tools. * Sanitization: None mentioned; the agent assumes the output is a legitimate log of errors and tests.
[External Downloads] (SAFE): Includes a link to documentation on GitHub (github.com/Falkicon/afd). This is used for reference and does not trigger automated downloads or execution of untrusted code.
[Safe] (SAFE): All other security categories, including Credentials, Obfuscation, and Persistence, showed no suspicious patterns.

s-work