The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted content such as business ideas, market evidence, and customer personas, creating an attack surface for indirect prompt injection.
Ingestion points: The workflow in SKILL.md explicitly requires external inputs, including "Business idea, offer, and target-customer assumptions" and "Existing market evidence, customer conversations, or draft claims."
Boundary markers: The instructions lack requirements for the use of delimiters (e.g., XML tags or specific markdown blocks) or "ignore embedded instructions" warnings to help the agent distinguish between its core validation logic and the data provided for analysis.
Capability inventory: The skill leverages the agent's core analytical capabilities to perform market audits and generate recommended revisions. While the skill files do not contain scripts for direct system or network access, the agent's ability to act on instructions within the analyzed data poses a risk of instruction override.
Sanitization: There are no mechanisms or instructions provided to validate, filter, or sanitize the external data before it is integrated into the agent's reasoning process.

meta-market-validation