law-to-markdown
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from user-provided legal documents, which are then presented to the agent in Markdown format.
- Ingestion points: Documents are read and parsed in
scripts/law_to_markdown.pyandscripts/cn_law_normalizer.pythrough libraries likepdfplumberandpython-docx. - Boundary markers: The skill does not implement specific delimiters or "ignore previous instructions" warnings when presenting the processed text to the agent.
- Capability inventory: The skill possesses significant capabilities, including executing subprocesses via
osascript(to call external OCR tools) and performing file system operations such as writing, copying, and deleting files. - Sanitization: While the skill normalizes the text for formatting and legal structure, it does not perform sanitization to escape or neutralize potential prompt injection instructions embedded within the source documents.
Audit Metadata