pp-medium-reader

Pass

Audited by Gen Agent Trust Hub on Aug 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installation involves fetching binaries and source code from the vendor's GitHub and NPM repositories (@mvanhorn/printing-press-library).
  • [DATA_EXFILTRATION]: The CLI tool provides a --deliver webhook:<url> flag, enabling the agent to POST results to arbitrary external URLs. Additionally, it includes a feedback feature that can transmit local logs to a remote endpoint if specific environment variables are configured.
  • [COMMAND_EXECUTION]: The skill requires the agent to install and execute the medium-reader-pp-cli binary to perform core functions.
  • [CREDENTIALS_UNSAFE]: The documentation describes methods for providing a sensitive MEDIUM_SESSION cookie via environment variables to bypass paywalls. It also mentions an optional build tag that could allow automatic browser cookie extraction.
  • [PROMPT_INJECTION]: The skill fetches and processes untrusted content from Medium articles and RSS feeds. Ingestion points: CLI output from read, feed, and search commands. Boundary markers: None identified. Capability inventory: Binary execution, network operations (Medium/webhooks), and local database storage. Sanitization: None described for external content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 17, 2026, 01:29 PM
Security Audit — agent-trust-hub — pp-medium-reader