The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted Flutter and Dart code, which introduces a surface for indirect prompt injection. Ingestion points: Read, Glob, and Grep tools are used to read file content (SKILL.md). Boundary markers: The instructions lack explicit delimiters or instructions to the agent to ignore embedded commands in reviewed code. Capability inventory: The skill is restricted to read-only tools including Read, Glob, Grep, and mcp__very-good-cli__packages_check_licenses (SKILL.md), with no write or network access granted in allowed-tools. Sanitization: No sanitization of ingested code content is described.
[EXTERNAL_DOWNLOADS]: The skill recommends using osv-scanner for dependency audits, pointing to a repository managed by a well-known organization (Google). Evidence: The skill references https://github.com/google/osv-scanner/releases in references/supply-chain.md as a source for the tool. This is a trusted reference for a standard security utility.

static-security