nano-pdf

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the nano-pdf package using pip or uv from the Python Package Index (PyPI), which is a well-known service for software distribution.
  • [COMMAND_EXECUTION]: The skill utilizes the nano-pdf CLI to perform edits on local files, which involves executing shell commands with arguments derived from user input or external data.
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves processing external PDF files and natural language instructions, creating a surface for potential indirect prompt injection attacks.
  • Ingestion points: The skill ingests untrusted data from local PDF files and processing instructions passed to the nano-pdf edit command in SKILL.md.
  • Boundary markers: There are no explicit boundary markers or delimiters mentioned to separate processed data from instructions, which could lead the underlying LLM to follow malicious instructions embedded within a PDF.
  • Capability inventory: The skill has the capability to modify the local file system through PDF editing and can perform package installations via pip.
  • Sanitization: There is no evidence of input validation, sanitization, or content filtering for the data read from the PDF files or the instructions provided to the CLI.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 07:01 PM
Security Audit — agent-trust-hub — nano-pdf