The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests and processes external code artifacts through specialized auditor agents. 1. Ingestion points: The skill reads file content using the Read, Glob, and Grep tools to populate the {conversation_context} and {scope_description} variables. 2. Boundary markers: The agent templates for the Security, Accessibility, and Code Standards auditors lack explicit delimiters or markers to isolate the audited content from instructions. 3. Capability inventory: The skill has the capability to spawn sub-agents via the Task tool, write reports via the Write tool, and perform limited filesystem operations using Bash. 4. Sanitization: No sanitization or content escaping is implemented before data is passed to the sub-agents.
[EXTERNAL_DOWNLOADS]: The skill documentation provides instructions for users to resolve dependencies by cloning from the author's official GitHub repository at github.com/mikeng-io/agent-skills.
[COMMAND_EXECUTION]: The skill is granted permission to execute specific Bash commands (git, ls, and mkdir) to verify the local environment and manage audit output directories.

deep-audit