The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local shell commands and scripts to process data.
In SKILL.md, the agent is instructed to use pdftotext to extract text from PDF files if direct reading fails.
The agent executes a local Python script scripts/export_excel.py via subprocess.run or command line to generate the final Excel output.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it processes untrusted external content.
Ingestion points: Content is ingested from user-provided PDF and Word files as described in SKILL.md (Step 1).
Boundary markers: The instructions do not specify any delimiters or warnings to the model to ignore potential instructions embedded within the paper text.
Capability inventory: The skill has the ability to execute shell commands and write files to the local filesystem through its associated Python script.
Sanitization: There is no mention of sanitizing, escaping, or validating the text extracted from papers before it is processed by the LLM.

paper-analyzer