Give your AI Agent a senior engineer's debugging brain. Runbook-driven incident investigation via MCP — evidence collection, decision rules, structured reports. No more hallucinated root causes.
-
Updated
Jun 21, 2026 - HTML
#
llm-debugging
Here are 11 public repositories matching this topic...
Root cause analysis for AI agents. Detects agent loops, retry storms, and optimization opportunities in LangSmith, Langfuse, Arize Phoenix, and OpenTelemetry traces.
phoenix multi-agent ai-agents arize root-cause-analysis opentelemetry ai-observability langsmith langfuse arize-phoenix llm-observability llm-debugging agent-debugging
-
Updated
Jul 5, 2026 - Python
Prompt and response tracing for LLM workflows
-
Updated
Mar 12, 2026 - Python
Forkline is a replay-first tracing and diffing library for agentic AI workflows that lets you deterministically reproduce, fork, and compare agent runs to find exactly where behavior diverged.
developer-tools ml-infrastructure cli-tools agentic-ai ai-infrastructure llm-debugging deterministic-replay trace-diffing
-
Updated
Feb 27, 2026 - Python
Zero-intrusion guard for LLM calls in dev: dedupe, cache, and protect AI requests across Node, browser, and Vite.
ai gemini-api token-cache non-intrusive openai-api ai-development-tools ai-application-development llm environment--development llm-streaming llm-api llm-application-development llm-api-development llm-development ai-streaming llm-debugging llm-api-optimizer ai-cost ai-infrastructure-development zero-intrusion
-
Updated
Apr 16, 2026 - TypeScript
OKI TRACE: Local LLM observability. See step-by-step, layer-by-layer what your AI thinks. Logit Lens & Attention for HuggingFace models.
python open-source ai transformers developer-tools attention-mechanism blackbox huggingface ai-tools mechanistic-interpretability local-llm ai-interpretability llm-observability ai-transparency glass-box-ai llm-debugging logit-lens
-
Updated
May 17, 2026 - Python
🧠 Local-first LLM state branching & debugging. Capture, visualize, branch, and regression-test LLM calls as a DAG. Free & local via Ollama, any OpenAI-compatible API, MCP for agents. Zero config.
encryption mcp regression-testing dag ai-agents local-first llm prompt-engineering openai-proxy ollama context-management llm-observability model-context-protocol agent-memory claude-code claude-plugin llm-debugging
-
Updated
Jul 8, 2026 - JavaScript
ContextForge Core 的本地优先（local-first）可视化控制台 —— 索引/检索/rerank/memory/citation/eval 全链路调试与可观测
go nextjs evaluation gin developer-tools memory-management observability rag local-first retrieval-augmented-generation contextforge llm-debugging
-
Updated
Jul 4, 2026 - Go
Stop guessing why your MLX model outputs garbage. Triage in 30 seconds — no model load required.
python macos cli quantization mlx model-debugging apple-silicon safetensors local-llm mlx-lm llm-debugging
-
Updated
Mar 13, 2026 - Python
Self-hosted LangGraph debugger for VS Code — step through nodes, diff state, time-travel, per-node tokens, and paste any chat (Copilot/ChatGPT) as the LLM.
vscode-extension react-flow langchain langgraph time-travel-debugging llm-debugging ai-agent-debugging langgraph-debugger langgraph-studio-alternative
-
Updated
Jul 7, 2026 - TypeScript
LLM Harness for developing, debugging, and evaluating custom agents, tools, MCP servers, prompts, and memory constructs using Anthropic, OpenAI, or a local model.
agent tools mcp agents llm llms llm-agent llm-inference llm-tools llm-evaluation agentic-ai mcp-servers mcp-server mcp-client mcp-tools agent-orchestration agent-skills llm-debugging mcp-debugger harness-engineering
-
Updated
May 21, 2026 - Python
Improve this page
Add a description, image, and links to the llm-debugging topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the llm-debugging topic, visit your repo's landing page and select "manage topics."