fulfill-git-escrow
Warn
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: MEDIUMDATA_EXFILTRATIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill requires the agent to verify the existence of a .env file and interact with it to manage a sensitive PRIVATE_KEY used for blockchain authentication and reward collection.
- [EXTERNAL_DOWNLOADS]: The skill depends on the installation of the git-escrows CLI from the NPM registry (npm i -g git-escrows), which is a package managed by the author arkhai-io and not a recognized trusted vendor.
- [COMMAND_EXECUTION]: The skill utilizes shell commands to clone source code from remote git repositories provided via user input or escrow metadata for the purpose of analysis and solution development.
- [PROMPT_INJECTION]: Processing external code from cloned repositories presents a surface for indirect prompt injection attacks. Ingestion points: Source code from remote repositories is cloned to a temporary directory and read by the agent in Step A1. Boundary markers: The instructions lack delimiters or specific directives to ignore potential instructions embedded within the downloaded test files. Capability inventory: The agent has access to powerful tools including Bash for command execution and Write/Edit for modifying local files. Sanitization: There is no evidence of validation or sanitization of the content from the external repositories before the agent analyzes it.
Audit Metadata