replacing-text

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE] (SAFE): The skill consists entirely of Markdown documentation (SKILL.md and reference/sd-guide.md). No executable scripts (.sh, .py, .js), binaries, or configuration files for automated execution are included.
  • [COMMAND_EXECUTION] (SAFE): Examples provided in the documentation use standard, well-known CLI tools (sd, fd, ripgrep, jq, git) for their intended purposes in a software development workflow. There are no signs of command obfuscation or suspicious shell patterns.
  • [PROMPT_INJECTION] (SAFE): The skill describes processing untrusted text data, which represents a surface for indirect prompt injection. Ingestion points: generic text files and command pipelines; Boundary markers: absent; Capability inventory: file replacement (sd) and version control (git); Sanitization: absent (standard CLI behavior).
  • [DATA_EXPOSURE] (SAFE): The documentation does not reference sensitive file paths, environment variables, or hardcoded credentials. All examples use generic filenames such as 'file.txt' and 'config.json'.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:11 PM