The Agent Skills Directory

[SAFE]: The skill operates as a workflow orchestrator for academic verification. It utilizes a separate skill (perplexity-research) to interact with established academic platforms and databases.
[EXTERNAL_DOWNLOADS]: The skill references several well-known academic services (Retraction Watch, PubPeer, OSF, Semantic Scholar, OpenAlex). These are recognized as legitimate research tools and do not involve the execution of untrusted code or scripts.
[COMMAND_EXECUTION]: The skill uses gbrain CLI commands for local knowledge management. These are standard operations for the platform environment and are restricted to querying and writing structured research data to specified directories (concepts/).
[INDIRECT_PROMPT_INJECTION]: The skill processes external research data but implements a structured verification flow with specific status levels (Verified, Partially verified, etc.). It uses these categories to prevent the agent from blindly trusting external content. The use of structured markdown templates for output provides clear boundaries for ingested data.

academic-verify