plugin-syncer

Warn

Audited by Gen Agent Trust Hub on Jun 12, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTIONNO_CODE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute a local Python script, scripts/sync_with_inventory.py, which orchestrates the environment synchronization process.
  • [EXTERNAL_DOWNLOADS]: The skill is explicitly designed to fetch skills and plugins from external GitHub repositories and local directories based on a provided inventory.
  • [REMOTE_CODE_EXECUTION]: By fetching and executing code from external sources defined in a configuration file, the skill performs remote code execution as its primary intended function.
  • [NO_CODE]: The script responsible for the synchronization logic, scripts/sync_with_inventory.py, is referenced in SKILL.md but is missing from the provided files. This prevents verification of safety measures such as URL validation or repository integrity checks.
  • [PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection by processing untrusted data to perform high-privilege actions.
  • Ingestion points: Reads repository URLs and plugin metadata from plugin-sources.json.
  • Boundary markers: None identified; there are no instructions to the agent to ignore malicious commands embedded in fetched plugin data.
  • Capability inventory: The skill has access to Bash, Read, and Write tools and executes code downloaded from the internet.
  • Sanitization: No evidence of URL validation, checksum verification, or source sanitization is provided in the documentation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 12, 2026, 08:07 PM
Security Audit — agent-trust-hub — plugin-syncer