The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources, which presents an inherent surface for indirect prompt injection attacks.
Ingestion points: The skill reads external data from free public pages such as a company's own site (as seen with quillhr.com in SKILL.md) and local workspace files such as PERSONA.md and ICP.md.
Boundary markers: No specific boundary markers or instructions to distinguish data from instructions are defined for the agent when processing external web content.
Capability inventory: The skill is restricted to reading files and making network requests for research; the Procedure and QC sections explicitly state the skill "writes nothing" and "no file changed."
Sanitization: No explicit sanitization, escaping, or filtering of external content is mentioned.
[EXTERNAL_DOWNLOADS]: The Exceptions section in SKILL.md provides a command for the user to execute (npx skills add eliasstravik/gtm-skills -s gtm-workspace -y) if a dependency is missing. This command installs a related skill from the same author (eliasstravik) and is part of the standard setup process.

gtm-qualify-prospects