VikingOwl91/gnoma

v0.3.4 Bugfix

This release fixes issues for SREs watching stability and regressions.

Published 10d AI Agents & Assistants

View tool

✓ No known CVEs patched

Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agentic ai-agent ai-coding-assistant anthropic cli coding-agent

+13 more

gemini go llama-cpp llm local-llm mcp mistral model-context-protocol multi-armed-bandit ollama openai self-hosted tui

Summary

AI summary

Fixes across provider/openai, router/discovery, and slm modules including request parameter wiring, capability probing, JSON enforcement, diagnostics improvements, timeout defaults, and error recovery.

Changes in this release

Type	Severity	Summary	CVE
Bugfix
Bugfix	Medium	Enforce JSON output and strip thinking-block prefixes in SL M. Enforce JSON output and strip thinking-block prefixes in SL M. Source: llm_adapter@2026-05-28 Confidence: high	—
Bugfix	Medium	Add honest classifier diagnostics and set default timeout to 15 seconds in SL M and router. Add honest classifier diagnostics and set default timeout to 15 seconds in SL M and router. Source: llm_adapter@2026-05-28 Confidence: high	—
Bugfix	Medium	Support classifier‑only SL M instances, add session error recovery, and provide feasibility diagnostics. Support classifier‑only SL M instances, add session error recovery, and provide feasibility diagnostics. Source: llm_adapter@2026-05-28 Confidence: high	—
Bugfix	Low	Always probe Ollama capabilities; cache usage is optional in router discovery. Always probe Ollama capabilities; cache usage is optional in router discovery. Source: llm_adapter@2026-05-28 Confidence: high	—
Bugfix	Low	Wire ResponseFormat into OpenAI request parameters for the provider/openai module. Wire ResponseFormat into OpenAI request parameters for the provider/openai module. Source: llm_adapter@2026-05-28 Confidence: high	—

Full changelog

Changelog

c065a2dea7732a869ea5d3ea7da18b1644ac8375 fix(provider/openai): wire ResponseFormat into OpenAI request params
fd327107df2bf7878febd6ceeef75eadb8e94782 fix(router/discovery): always probe ollama capabilities, cache is optional
c0c2e4bff572c1923f536e1abfbac564fad67abf fix(slm): enforce JSON output + strip thinking-block prefixes
f3c70bd802f693e35abd1ca6b3f11aee16cdbd66 fix(slm,router): honest classifier diagnostics + 15s default timeout
0d3d190a8b83a00975abc4c7e9d40a43e84839b1 fix(slm,session,router): classifier-only SLMs + session error recovery + feasibility diagnostics

View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track VikingOwl91/gnoma

Get notified when new releases ship.

About VikingOwl91/gnoma

All releases →