aigo-client-collab

Pass

Audited by Gen Agent Trust Hub on Oct 27, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill implements an automated self-update mechanism in scripts/check_update.py. This script periodically connects to the vendor's GitHub repository to compare version numbers and download updated content.
  • Evidence: Fetches version information and archive data from https://raw.githubusercontent.com/AI-GO-APP/aigo-client-collab-skill/main/VERSION and https://github.com/AI-GO-APP/aigo-client-collab-skill/archive/refs/heads/main.zip.
  • [COMMAND_EXECUTION]: The skill instructions direct the AI agent to execute various shell commands to manage the development environment, version control, and skill installation.
  • Evidence: Uses commands such as gh (GitHub CLI) for issue management and authentication checks, git for repository cloning and updates, and python to run internal utility scripts.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external sources, specifically GitHub issues and client-provided text, which represents a potential attack surface for indirect prompt injection.
  • Ingestion points: Data is pulled into the agent's context via gh issue view and from user prompts containing feature requests or bug reports.
  • Boundary markers: The skill uses structured templates for issues and research tasks (defined in references/issue-format.md) to maintain clear data boundaries.
  • Capability inventory: The agent has access to file system operations (via git), issue management (via gh), and local script execution (python).
  • Sanitization: Instructions strictly prohibit the agent from including PII, credentials, or keys in outputs and require explicit user review and confirmation before the agent submits any content back to external platforms.
  • [DYNAMIC_EXECUTION]: The PDF generation utility pdf/guide.py employs dynamic loading to import components from external paths specified in configuration files.
  • Evidence: The script modifies the Python sys.path at runtime using a path provided in client.json or an environment variable to import the Deck class from a separate presentation skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 27, 2026, 06:48 AM