essay-publishing-pipeline

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides example pandoc commands for converting markdown to HTML and PDF. These are standard document processing utilities and do not involve untrusted input or risky execution patterns.- [DATA_EXFILTRATION]: No exfiltration patterns were found. The provided Python snippets for RSS generation and cross-posting use structured metadata and target user-controlled API endpoints (e.g., Dev.to, Newsletters) as part of the intended publishing workflow.- [PROMPT_INJECTION]: The documentation does not contain instructions that attempt to override agent behavior or bypass safety filters. It uses natural instructional language to describe a publishing workflow.- [EXTERNAL_DOWNLOADS]: The skill does not perform any remote code downloads or external script executions. The Python code uses standard libraries (pathlib, yaml, markdown, xml.etree.ElementTree) for local file processing.- [SAFE]: The Python script correctly uses yaml.safe_load(frontmatter) to prevent unsafe deserialization vulnerabilities when parsing essay metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 06:08 AM
Security Audit — agent-trust-hub — essay-publishing-pipeline