The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests untrusted user-supplied data to generate complex business documents and interacts with file-writing tools.
Ingestion points: User-provided inputs for {idea}, {target_market}, {team_info}, and {initial_capital} in SKILL.md are interpolated directly into the workflow.
Boundary markers: The skill lacks explicit boundary markers or instructions to the agent to disregard potential instructions embedded within the user-provided startup data.
Capability inventory: The skill has the capability to chain with moai-officer:doc-xlsx and moai-officer:doc-pptx to generate and save documents based on the processed input.
Sanitization: There is no evidence of input validation or sanitization logic to filter out adversarial instructions within the provided variables.

consult-startup