The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external CRM and intelligence platforms, creating an attack surface for indirect prompt injection where malicious data could influence agent behavior.\n
Ingestion points: External data is ingested via MCP tools such as hubspot_get_deal (deal notes), clari_get_call_summary (discovery pain quotes and objections), and apollo_organizations_enrich (company enrichment data) in SKILL.md.\n
Boundary markers: The skill lacks explicit boundary markers or delimiters when interpolating this external data into the generated RECAP framework or the post-demo scorecard template.\n
Capability inventory: The skill possesses impactful write capabilities, including hubspot_create_contact_note and hubspot_create_contact_task for CRM updates, as well as gmail.create_draft for automated follow-ups.\n
Sanitization: There is no evidence of sanitization, escaping, or validation of the external content before it is used to generate outputs or written back to the HubSpot CRM.

demo-execution-playbook-skill