@cloudflare/codemode@0.5.2
charmbracelet/gum@cloudflare/codemode@0.5.2Sep 11, 2026by github-actions[bot]
Upgrading?
AI Summary
This patch release introduces structural truncation for oversized data structures to prevent models from receiving incomplete JSON documents. It refines the handling of large results by shrinking values in place and bounds model context by projecting calls and limiting sandbox logs. The changes also apply response budgets to MCP server text output.
Key Highlights
- Truncates structured results structurally instead of slicing JSON strings to ensure valid output.
- Oversized values are shrunk in place, prioritizing the largest elements first.
- Code Mode tools now use a `toModelOutput` projection to manage model context.
- Durable call logs are separated from model context to prevent overflow.
- Fixes issues related to oversized objects and logs.
New Features
- Structural truncation logic for strings, arrays, and objects
- New `toModelOutput` projection for Code Mode tools
- Response budget application for MCP servers
Full Release Notes
### Patch Changes
- [#2224](https://github.com/cloudflare/agents/pull/2224) [`dcca089`](https://github.com/cloudflare/agents/commit/dcca0896cc829e4a602a1feb47466ac79d1ca096) Thanks [@mattzcarey](https://github.com/mattzcarey)! - Truncate structured results structurally instead of slicing their JSON.
`truncateResult` (the default `transformResult` in Think's execute tool and the browser tools) and the `codeMcpServer` / `openApiMcpServer` response path used to pretty-print an oversized object and cut it at a character count, leaving the model with half a JSON document. Oversized values are now shrunk in place, largest values first: strings are clipped with a `--- TRUNCATED --- <n> chars` suffix, arrays keep their leading items and end with a `--- TRUNCATED --- <n> more items` element, and an object only loses entries (largest first, named in a `"--- TRUNCATED ---"` entry) when its values cannot share the budget. The output is always valid JSON of the original shape and small values pass through untouched.
The Code Mode tool now carries a `toModelOutput` that projects `calls` out of what the model sees and bounds the sandbox `logs` the same way a result is bounded. The durable call log — every connector call's args and result — stays on the persisted tool part for UIs and audit, but no longer rides into the model's context uncapped alongside the transformed `result`. The projection never throws: a BigInt or cyclic value is rendered rather than failing a completed run at model assembly.
The MCP servers apply the response budget to the text they emit, pretty-printing only while that still fits.
Fixes [#2009](https://github.com/cloudflare/agents/issues/2009), [#2143](https://github.com/cloudflare/agents/issues/2143) and [#2182](https://github.com/cloudflare/agents/issues/2182).
Related Repositories
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
A fast reverse proxy to help you expose a local server behind a NAT or firewall to the internet.
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context layer for LLMs
The world’s fastest framework for building websites.
:cherry_blossom: A command-line fuzzy finder