orwell-writing

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to ingest and revise untrusted prose, creating an indirect prompt injection surface where malicious instructions could be embedded in the content being edited.
  • Ingestion points: User-provided drafts, essays, and documentation as specified in SKILL.md.
  • Boundary markers: Absent; the skill does not define structural delimiters to isolate untrusted input.
  • Capability inventory: The skill inherits the agent's full capabilities (e.g., file system access), as no tool restrictions are defined in the frontmatter.
  • Sanitization: No filtering or sanitization of input text is specified.
  • [EXTERNAL_DOWNLOADS]: The README.md file references an installation command using an external NPM package (@tamng0905/builder-essential-skills) which is not from a trusted organization or the explicitly named skill author.
  • [NO_CODE]: This skill consists entirely of instructional Markdown and YAML configuration files, containing no executable scripts or binaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 01:23 PM
Security Audit — agent-trust-hub — orwell-writing