The Agent Skills Directory

[SAFE]: The skill is a collection of Markdown instructions and templates designed to guide an agent through a research workflow. It does not contain any executable scripts, binaries, or automated installation procedures.
[INDIRECT_PROMPT_INJECTION]: As a research-focused skill, it is designed to ingest and process data from external sources (web, documents, APIs). While this presents a surface for indirect prompt injection, the skill specifically mandates a rigorous evaluation process using the CRAAP test (Currency, Relevance, Authority, Accuracy, Purpose) and triangulation across multiple independent sources to mitigate the risk of acting on malicious or inaccurate data.
[COMMAND_EXECUTION]: The technical verification track (references/technical-verification.md) suggests using command-line tools such as curl, python, and nvidia-smi to reproduce and verify technical claims. These instructions are presented as standard research procedures and are not malicious in intent or execution.
[EXTERNAL_DOWNLOADS]: The skill mentions using retrieval tools to gather research material. These activities are limited to the primary function of the skill (research) and do not involve downloading or executing unverifiable code from untrusted remote servers.

research-methodology