pdfcraft-redact-pdf

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The configuration file references/craft-command-evidence.json references a download link for a Windows portable executable (pdfcraft-cli.exe) hosted on GitHub at https://github.com/storytold/pdfcraft/releases/download/v0.8.0/pdfcraft-0.8.0-windows-x64-portable.zip. This is provided as technical evidence of the runtime version used by the toolset.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process PDF documents provided by users. This represents a standard attack surface where malicious text or instructions embedded in a PDF could potentially target the AI agent. The skill implements safety measures by instructing the agent to perform multiple verification steps (searching, copying, and extracting) to confirm that sensitive content has been permanently removed.
  • [COMMAND_EXECUTION]: The documentation in references/craft-methods.md and SKILL.md describes the execution of CLI commands and MCP tool calls for document manipulation, such as doc_open, redact_apply, and doc_save. These are documented as standard operations for the PDF redaction workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 06:47 PM