Repository Issue Activity (beta)

#1517 State of the project

#1540 the agent's instructions can reference the role settings

#1546 docs: hosted openai-compatible client example (name:model id)

#1542 Bedrock: tool results with non-object JSON output are rejected (toolResult.content[].json must be an object)

#1541 Docs: add AnyLLM as a documented OpenAI-compatible provider (24+ models, one key)

#1539 Add deepseek-v4-flash / deepseek-v4-pro model support

#1241 I don't see how to prevent think block

#1537 Add Eden AI as a provider

#1424 Support Ollama’s Enable / Disable Thinking

#1536 💡 [Call for Contributions] Share Your Creative aichat Use Cases & Workflows!

#1523 Suddenly no memory.

#1528 Qwen3.6 thinking mode without <think> tag

#1527 Use $EDITOR as default editor if set

#1489 execute flag doesn't give executable commands

#1524 Clarify local agent setup: config.yaml alone is not sufficient

#1519 Update model file to add Claude Opus 4.7 versions

#1514 support DeepSeek-v4 series model

#801 Any method to visualize markdown contents?

#1506 How to enable Gemini's google_search tool?

#1501 Suggestion: Support memvid instead of sparse rag files

#1457 Feature request: re-run init wizard / interactive model picker for existing config (sync from /v1/models)

#1490 Implement Agent Client Protocol (ACP)?

#1504 Port apfel demos into aichat

#1475 Remove hard coded color in shell exec

#1476 Aichat --info shows info before config.yaml cmd_prelude