The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use shell commands for verification, specifically using curl and grep to check the status of URLs. This involves executing network-enabled commands that process external data.
Evidence: curl -sIL https://directory.com/your-listing | grep -i rel= in SKILL.md and references/directory-list.md.
[EXTERNAL_DOWNLOADS]: The directory reference list includes a URL, https://fitprofessionals.net, which is currently flagged as malicious by security scanners. Recommending users interact with or submit data to a blacklisted domain poses a security risk.
Evidence: Found in references/directory-list.md and references/submission-tracker-template.csv.
[INDIRECT_PROMPT_INJECTION]: The skill is configured to read product marketing context from local files (e.g., .agents/product-marketing.md) without implementing sanitization or boundary markers. This creates an attack surface where malicious instructions in these files could exploit the agent's command execution capabilities.
Ingestion points: SKILL.md instructs reading .agents/product-marketing.md, .claude/product-marketing.md, and product-marketing-context.md.
Boundary markers: None provided in the instructions.
Capability inventory: Shell command execution via curl and grep.
Sanitization: None described for external context data.
[SAFE]: The skill references many reputable and well-known services like Product Hunt, G2, Capterra, GitHub, and various official integration marketplaces, which are standard for the described use case.

directory-submissions