sauve-jazz-extension

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill's README.md and install.sh reference a remote installation script and a git repository hosted at https://raw.githubusercontent.com/Alcyone-Labs/skill-forge/. These are vendor-owned resources used for the legitimate purpose of installing and updating the skill files.
  • [REMOTE_CODE_EXECUTION]: The installation process involves downloading a shell script (install.sh) from the official repository of the skill author (Alcyone-Labs) and piping it directly to the bash interpreter. This is a common deployment pattern for developer tools within this ecosystem and is verified as a vendor resource.
  • [COMMAND_EXECUTION]: The install.sh script executes system commands including mkdir, rm, cp, and git clone to manage the local environment and install skill files into the agent's configuration directories (e.g., ~/.config/opencode/skills).
  • [PROMPT_INJECTION]: As the skill is designed to process external data sources like RSS feeds and bookmarks, it possesses an indirect prompt injection surface. However, the documentation emphasizes the use of a structured JSON message protocol and serialization logic (serializeCoValue) which provides mitigation by establishing data boundaries between the untrusted content and the agent's core logic.
  • [SAFE]: No evidence of malicious behavior such as data exfiltration, hardcoded credentials, or code obfuscation was detected. All network and file system operations are consistent with the skill's primary purpose of providing developer documentation and installation tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 06:42 PM