The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest findings from other audit tools and "carry forward" verification commands into a remediation plan. If an upstream audit source is compromised or contains malicious instructions, the agent may incorporate those instructions or harmful shell commands into its output for the user to execute.
Ingestion points: Audit findings and reference files (e.g., robots-ai-crawler-audit/references/checks.md) from external search/SEO audit skills as described in SKILL.md and references/checks.md.
Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded prompts within the findings it processes.
Capability inventory: The skill is designed to provide users with executable bash commands (such as curl and grep) to verify fixes.
Sanitization: There are no instructions provided to sanitize or validate the commands or text strings carried forward from the source audits.

ai-search-remediation-plan