deploy-to-mainnet

Pass

Audited by Gen Agent Trust Hub on Aug 1, 2026

Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill implements a telemetry system that transmits system metadata to an external endpoint.
  • Evidence: Preamble scripts in SKILL.md extract operating system and machine architecture information (uname) and transmit it via curl to a URL retrieved from the user local configuration file (~/.superstack/config.json).
  • Context: While the skill prompts for consent, it transmits system-level identifiers to an external domain that is not on the whitelisted services list.
  • [COMMAND_EXECUTION]: The skill uses multiple bash script blocks for operational logic and framework management.
  • Evidence: SKILL.md contains shell commands for reading/writing configuration files, generating timestamps, and executing Solana-specific deployment tools like anchor and solana CLIs.
  • Risk: Extensive shell use for maintenance increases the operational surface area.
  • [PROMPT_INJECTION]: The skill processes external data that could contain instructions intended to influence agent behavior.
  • Ingestion points: The main workflow in SKILL.md reads .superstack/build-context.md to verify previous build and test phases.
  • Boundary markers: Absent. The instructions do not include delimiters or specific warnings to the agent regarding potential instructions embedded within the context file.
  • Capability inventory: The skill utilizes shell execution (bash) and performs file system operations and network requests via CLI tools.
  • Sanitization: Absent. The content from the external file is used directly to determine the workflow path without validation or escaping.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 1, 2026, 11:05 AM
Security Audit — agent-trust-hub — deploy-to-mainnet