Skills
skills/laiye-ai/adp-cli/agentic-doc-parse-and-extract/Gen Agent Trust Hub

agentic-doc-parse-and-extract

Fail

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches installation scripts and pre-compiled binaries from the vendor's official GitHub repository and cloud storage endpoints (aliyuncs.com).
  • [REMOTE_CODE_EXECUTION]: Provides instructions to download and execute setup scripts (adp-init.sh and adp-init.ps1) directly from the vendor's GitHub repository.
  • [COMMAND_EXECUTION]: Executes the adp command-line tool to perform document parsing, extraction, and configuration. The installation guide includes steps that modify system environment variables and suggest the use of elevated privileges via sudo or administrator-level commands.
  • [PROMPT_INJECTION]: The skill processes document content from external URLs and local paths, which constitutes an indirect prompt injection surface.
  • Ingestion points: Use of adp parse url, adp extract url, and local folder processing commands in SKILL.md and references/commands.md.
  • Boundary markers: The skill does not instruct the agent to use delimiters or ignore instructions within the documents being processed.
  • Capability inventory: The agent has access to shell command execution (adp), network operations, and file read/write capabilities through the CLI tool.
  • Sanitization: No explicit sanitization or validation of the processed document content is documented before it is returned to the agent context.
Recommendations
  • HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/laiye-ai/adp-cli/main/scripts/adp-init.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Apr 30, 2026, 10:48 AM