The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to ingest and process data from external web sources. Specifically, modules/research-checklist.md instructs the agent to perform web research and update its core guidance based on the findings. * Ingestion points: Web research tools (distill_from_wikipedia, distill_from_arxiv, and general web search) specified in modules/research-checklist.md. * Boundary markers: Absent. There are no instructions to the agent to treat external content as untrusted or to use delimiters. * Capability inventory: The agent is instructed to perform deep scholarly analysis and update its own internal modules based on the research. * Sanitization: Absent. The workflow does not include steps to filter or validate the ingested content for malicious instructions.
[EXTERNAL_DOWNLOADS]: The modules/research-checklist.md file suggests using tools to fetch content from well-known academic and encyclopedic sources such as Arxiv and Wikipedia to validate and update legacy guidance.
[NO_CODE]: The skill consists entirely of instructional Markdown files and configuration and does not contain any executable scripts, binaries, or logic beyond natural language instructions.

biblical-exegesis-expert