Changelog

All notable changes to the DeepNerd platform (Vault IDE, AI Agents, DeepNerd API) will be documented here.

[v1.2.0] - 2026-04-20

Added

• Vault IDE: Introducing new multi-agent debugging sessions.
• DeepNerd API: New /v1/models/fine-tune endpoint for custom LoRA tuning.
• Agents: Added cross-region file system sync capabilities.

Changed

• DeepNerd API: Rate limits for free tier increased from 100 req/min to 500 req/min.
• SDK: Python SDK now auto-retries on 429 status codes by default.

Fixed

• Vault IDE: Resolved an issue where the integrated terminal would lose context on reload.
• Agents: Fixed a memory leakage issue during long-running automation tasks.

[v1.1.0] - 2026-03-15

Added

• Vault IDE: WebAssembly (WASM) execution environment natively inside the editor.
• DeepNerd API: Support for streaming reasoning traces (chain-of-thought) in completion endpoints.

Changed

• General: Rebranded Codex Assistants to DeepNerd Agents across the ecosystem.

Security

• DeepNerd API: Enforced TLS 1.3 minimum requirement for all API interactions.

[v1.0.0] - 2026-01-01

Added

• Initial public release of the DeepNerd Platform.
• Vault IDE with local intelligence routing.
• DeepNerd API supporting /chat, /code, and /agents endpoints.
• JavaScript and Python official SDKs.
• Autonomous coding agent CLI workflows.