The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and analyze untrusted code from a project repository to generate a security report. This creates an attack surface where malicious instructions could be embedded within the project's source code (e.g., in comments or large string literals) to influence the agent's behavior or misrepresent the scan results.
Ingestion points: Untrusted source code files identified in Step 1 (via git diff or find) and subsequently read by tools in Step 2 and Step 3.
Boundary markers: Absent. The skill does not provide instructions to the agent on how to distinguish project content from malicious instructions intended for the agent.
Capability inventory: The skill performs shell command execution (Step 2/3), file system reads, and generates a formatted report (Step 5).
Sanitization: Absent. There is no mention of filtering or escaping project content before it is processed or included in the output report.
[EXTERNAL_DOWNLOADS]: The skill uses 'npx' to run several industry-standard security scanners, which may involve downloading these tools from the official npm registry.
Evidence: Instructions to run npx snyk test, npx eslint, and npx @semgrep/semgrep in SKILL.md.
Context: These tools are from well-known security organizations and are used here for their intended purpose.
[COMMAND_EXECUTION]: The skill relies on executing various shell commands to automate the security auditing process.
Evidence: Uses git, find, npm audit, gitleaks, and grep to analyze the local project files and dependencies.

detecting-security-vulnerabilities