Closing the execution gap, Part 2: Dependency management

Short summary

jhansi.io v0.3 fixes AI-code friction by caching dependencies in a persistent workspace. Manifests (pyproject.toml, requirements.txt) or pipreqs auto-detect what to install; cold start drops from 40s to instant on repeat executions. Supports Python, Node, Go, Java with isolation per sandbox and restricted egress to official registries.

• •Persistent workspace caches dependencies for reuse across sandbox executions
• •Supports manifest files (pyproject.toml, requirements.txt) or pipreqs auto-detection
• •Reduces cold start time from 40s+ to instant for repeat runs; critical for AI agent performance

Generated with AI, which can make mistakes.