perplexity-upgrade-migration

Pass

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [DYNAMIC_CONTEXT_INJECTION]: The skill uses dynamic execution blocks to verify the presence and version of the OpenAI SDK in the user's environment before providing migration steps.
  • [COMMAND_EXECUTION]: Employs standard development tools such as grep and git to identify deprecated API patterns in local source code and manage upgrade branches.
  • [EXTERNAL_DOWNLOADS]: Facilitates upgrading to the latest official openai package via standard package managers (npm and pip).
  • [REMOTE_CODE_EXECUTION]: Includes a validation script that uses node -e to verify connectivity with the Perplexity API. This script targets the official canonical URL (https://api.perplexity.ai) and uses standard environment variables for authentication.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes local codebase files to find legacy strings. While this represents an ingestion point for untrusted data, it is limited to identifying specific architectural patterns for migration purposes.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 21, 2026, 09:39 PM
Security Audit — agent-trust-hub — perplexity-upgrade-migration