writing-for-humans

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill suggests installing an additional skill named 'humanizer' from the author's repository using the command npx skills@latest add jd-solanki/skills --skill humanizer. This targets a vendor-owned resource associated with the skill's author.\n- [COMMAND_EXECUTION]: Provides a shell command snippet intended for the user or agent to execute to expand the agent's capabilities by downloading and adding new skills.\n- [INDIRECT_PROMPT_INJECTION]: The skill guides the agent in processing user-supplied drafts, which creates a surface for potential indirect prompt injection.\n
  • Ingestion points: User-provided drafts for tickets, pull request bodies, and commit messages (SKILL.md).\n
  • Boundary markers: No specific delimiters are defined to separate user content from the skill's instructions.\n
  • Capability inventory: This skill is primarily instructional markdown and does not perform direct file system or network operations itself.\n
  • Sanitization: The instructions proactively include a security check, advising the user to redact API keys, tokens, customer data, and credentials from output before publishing.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 06:49 AM
Security Audit — agent-trust-hub — writing-for-humans