aps
Pass
Audited by Gen Agent Trust Hub on Apr 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
apscommand-line utility for all operations, including searching backends, managing local libraries, and opening PDFs in system-default viewers. - [EXTERNAL_DOWNLOADS]: Fetches academic paper metadata and PDF documents from well-known repositories such as Semantic Scholar and OpenAlex. It also supports user-configured sources for PDF resolution via
aps lib config. - [CREDENTIALS_UNSAFE]: Provides a mechanism (
aps login) to store API keys for Semantic Scholar and OpenAlex in a local configuration directory (~/.config/aps/), which is a standard approach for CLI-based authentication management. - [PROMPT_INJECTION]: The skill has an indirect injection surface as it ingests untrusted content from academic papers and manifest files. Ingestion points: Paper text is read via
aps lib readand discovery jobs are processed from manifest files viaaps scan. Boundary markers: None specified. Capability inventory:apsCLI operations for file management and system viewing. Sanitization: None described.
Audit Metadata