trace-analysis

BitDive Model Context Protocol (MCP) server. The Autonomous Quality Loop for AI agents. Provides real runtime context, before/after trace comparison, and integration testing workflows.

ARM64 trace evidence analysis & cipher algorithm recovery — Claude Desktop plugin with skills + local MCP server driving the native ak_search engine over GB-scale trace files

Utilities for analyzing trace data from ROS 2 systems generated by the ros2_tracing packages.

A governed learning layer for AI agents — turns execution traces into reviewed memories, reusable skills, and evidence-backed training data.

Local-first review infrastructure for smart contract security.

Analysis engine inside QuickCall — ingests AI coding session traces, runs multi-stage LLM pipeline, surfaces root causes and recurring failure patterns

Langfuse MCP server with built-in analytics. 34 tools — traces, observations, sessions, scores, prompts, datasets + accuracy metrics, failure detection, token percentiles, cost breakdowns, latency analysis, context breach scanning. Works with Claude Code, Cursor, Codex.

Local trace-import lab for estimating AI agent workflow cost, retry risk, pricing sensitivity, and cost per successful run before launch.

Nebula-Copilot 是一个面向值班与排障场景的智能 CLI 工具，支持 Trace 诊断、瓶颈定位、异常归因与可执行建议输出，并结合了 Agent Tool Calling 与知识库增强排障。

CLI tool for analyzing Playwright trace files without the browser-based viewer

Deterministic evaluation harness for tool-using AI agents: canonical traces, fault injection, policy scoring, and evidence with negative controls.

RAG-powered snow-sports Q&A with cited answers, hierarchical retrieval, retrieval sweeps, trace analysis, eval tooling, and a Gradio UI.

Local-first graph analytics for recurring paths, loops, failures, and outcomes across AI-agent traces and event data.

Trace-first release gate for coding-agent skills. Scores baseline vs. candidate traces across a 7-dimension rubric and emits a PASS / HOLD / INVESTIGATE Skill Delta Report.

Replay, regression, trace packaging, failure analysis, and dataset slicing for LLM agents

Out-of-tree HCCL Task DAG and P2P traffic capture studio

Agent trajectory forensics: loop detection, causal edges, drift, and failure-risk reports from local traces.

An eval and observability cockpit for coding agents. It runs policy-controlled coding agents in sandboxed toy repos, tool-use traces, MCP tools, compares harness policies, scores recovery and safety behavior with Python evals, and supports CI-gated verification and Braintrust/Weave export.