proof-writer

Pass

Audited by Gen Agent Trust Hub on Oct 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's primary function is text and mathematical generation. It reads from local project files (notes, theorem drafts) and writes to a designated documentation file (PROOF_PACKAGE.md). These operations are consistent with its stated purpose and do not involve sensitive system directories or unauthorized file access.
  • [SAFE]: There are no network-related commands, such as curl or wget, and no attempts to exfiltrate data to external domains. The environment is scoped to the local project root.
  • [SAFE]: The skill mentions an optional integration with a sibling skill (/lean-formalize), which is a modular pattern for formal proof verification. This invocation does not constitute arbitrary remote code execution as it refers to a local skill defined within the same workspace/author context.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from user-provided notes and sketches. While this constitutes an attack surface, the skill's capabilities are limited to writing Markdown files and performing mathematical reasoning. The instructions include a 'Feasibility Triage' step that mandates verification and explicit identification of missing assumptions, which acts as a safeguard against malicious or erroneous inputs influencing the agent's logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 24, 2026, 09:43 AM