skills/aeonfun/aeon/rightstack/Gen Agent Trust Hub

rightstack

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes npx to run a specific pinned version of the rightstack CLI (rightstack@0.3.2). The execution is handled via spawnSync without shell interpolation, and the environment passed to the child process is restricted to a whitelist of safe runtime variables. The skill is explicitly configured as read-only and prohibits any mutation of the repository or external systems.
  • [EXTERNAL_DOWNLOADS]: The skill uses npx --package rightstack@0.3.2 to ensure the tool is available. While this involves fetching the package if not cached, it targets a specific version of a relevant development tool consistent with the skill's stated purpose as a Web3 stack advisor. The implementation disables npm install scripts (npm_config_ignore_scripts: "true") to mitigate risks associated with post-install hooks.
  • [DATA_EXFILTRATION]: The skill includes a notification mechanism (./notify -f <path>), but this is used for local project reporting. The skill instructions explicitly prohibit sending sensitive data or executing commands found in the tool's output.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided build goals and tool names. It includes multiple validation steps, such as checking for schema versions in the tool's JSON output and requiring human-in-the-loop review of all recommendations to verify security and vendor-lock-in tradeoffs before any use.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 06:38 PM
Security Audit — agent-trust-hub — rightstack