OpenCode v1.15.13 Adds Session Metadata API, Fixes Anthropic Opus 4.7 Adaptive Reasoning Bug
OpenCode v1.15.13 introduces custom session metadata storage through its API and SDK. The release fixes a bug where Anthropic's Opus 4.7+ adaptive reasoning returned empty thinking blocks instead of summarized thinking.
OpenCode v1.15.13 Adds Session Metadata API, Fixes Anthropic Opus 4.7 Adaptive Reasoning Bug
OpenCode released version 1.15.13 with session metadata storage capabilities and a critical fix for Anthropic's Opus 4.7+ models.
Core Updates
The release adds custom metadata storage for sessions through OpenCode's API and SDK, contributed by community member @shantur. This allows developers to attach and retrieve custom data to coding sessions programmatically.
A notable bugfix addresses Anthropic's Claude Opus 4.7 and later versions when using adaptive reasoning mode. Previously, the gateway returned empty thinking blocks; the update now preserves summarized thinking content instead.
Configuration loading behavior changed to scan upward from the opened location. According to OpenCode, this makes directory-specific settings and provider policies apply more predictably across different project structures.
TUI Improvements
The terminal user interface received two bugfixes. Wrapped inline tool rows now maintain proper alignment, and failed inline tools can expand error details directly in place rather than requiring separate views.
What This Means
The session metadata feature gives developers structured storage for context, user preferences, or debugging information tied to specific coding sessions. This could enable more sophisticated workflow tracking and personalization in AI-assisted development tools.
The Anthropic Opus fix is particularly relevant for teams using Claude's latest models with chain-of-thought reasoning. Empty thinking blocks would have prevented visibility into the model's reasoning process, which is a key feature for understanding and debugging AI-generated code suggestions. The fix ensures summarized thinking remains accessible even when full reasoning chains are condensed.
Related Articles
Anthropic Python SDK v1.13.0 adds Managed Agents workflows, multiagent config and thread status filtering
Anthropic released v1.13.0 of its Python SDK, adding workflows, multiagent configuration and thread status filtering to Managed Agents. The release also adds types for the Chat and Cowork unified analytics metrics API. No new model is included.
Anthropic Python SDK 1.12.0 adds claude-haiku-5-5 and typed computer and browser tool calls
Anthropic's Python SDK v1.12.0, dated 2026-10-07, adds the claude-haiku-5-5 model identifier and typed tool calls for the computer and browser toolsets. It also adds lifecycle fields to /v1/models and several admin API changes. The release notes give no pricing, context window, or benchmark data for the new model.
Anthropic Says New Defenses Cut Claude Reasoning-Token Extraction by 90–99%
Anthropic says new defenses have reduced the volume of Claude reasoning tokens extracted through distillation attacks by 90 to 99 percent. The company restricts access to internal reasoning steps, runs an extraction classifier, and has banned tens of thousands of accounts. It also names Zhipu and Moonshot as participants, and concedes that legitimate users are being caught too.
Anthropic study: GLM-5.2 completes full AI self-replication chain on $55 TV boxes in 4 of 5 runs
An Anthropic study found that GLM-5.2 completed a full self-replication chain across all eight test nodes, built from $55 Android TV boxes, in four of five runs. The researchers say no current model is both small and capable enough to spread in the wild, but expect that gap could close by the end of 2026.
Comments
Loading...