The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends the installation of the deepxiv-sdk Python package. This is a vendor-owned dependency required for the skill to function when local scripts are unavailable.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute the deepxiv CLI and local Python adapter scripts (tools/deepxiv_fetch.py, tools/research_wiki.py). These commands incorporate user-supplied input from $ARGUMENTS, which requires careful handling by the agent to prevent shell command injection.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes content from external academic papers. Malicious instructions embedded in a paper's text could attempt to manipulate the agent's behavior during summarization or documentation tasks.
Ingestion points: Data enters the context from external academic papers via CLI stdout and script output (JSON/Text) in SKILL.md (Step 3).
Boundary markers: The instructions do not specify the use of delimiters or warnings to ignore instructions found within the retrieved paper content.
Capability inventory: The skill has access to Bash, Read, and Write tools, including the ability to write to a local research-wiki/ directory.
Sanitization: There is no explicit sanitization or filtering logic described for the content fetched from the DeepXiv service before it is presented to the user or saved to disk.

deepxiv