The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to install the axiv command-line tool using uv tool install axiv. The package is a vendor-owned resource provided by the skill author for the intended functionality of the skill.
[COMMAND_EXECUTION]: The skill uses the axiv CLI to manage research workflows and library organization. It enforces security boundaries by requiring explicit user authorization before executing state-changing operations, such as folder creation or deletion, using the --yes flag.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it ingests and processes content from external research papers and GitHub repositories.
Ingestion points: External data enters the context via axiv paper content (text extraction), axiv paper query (PDF querying), and axiv paper code (reading files from GitHub).
Boundary markers: The instructions do not specify the use of clear delimiters or warnings to isolate ingested content from the agent's primary instructions.
Capability inventory: The skill allows for local shell execution of the axiv CLI tool, including authenticated write operations to the user's alphaXiv library.
Sanitization: There is no evidence of content sanitization or validation performed on the text extracted from papers or code repositories before it is processed by the agent.
[PRIVILEGE_ESCALATION]: The installation guidelines explicitly advise against using sudo, ensuring the tool operates within the user's standard permission scope.

using-axiv-cli