The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from multiple external sources, which could potentially contain malicious instructions intended to influence the agent's behavior.
Ingestion points: Fields defined in references/form-schema.json include targetAccountInfo (pasted text), accountSourceUrl (external URL), and accountResearchFile (uploaded documents like PDF/Docx).
Boundary markers: No explicit delimiters or instructions are provided to the model to ignore potential prompt injection attempts within the analyzed content.
Capability inventory: The skill's primary function is text analysis and generation of summaries and action plans. The associated runner script (scripts/run.py) performs network operations to the vendor's API but does not provide the agent with local execution capabilities.
Sanitization: There is no evidence of sanitization or filtering of the external input before it is processed by the AI.
[EXTERNAL_DOWNLOADS]: The skill involves network interactions with external services.
Vendor API: The runner script scripts/run.py communicates with https://ai-skills.ai to process the skill's logic. This is the official infrastructure for the skill author.
User-Provided URLs: The skill can fetch and analyze content from URLs provided in the accountSourceUrl parameter, which is part of its core research functionality.

account-research-sales-card