[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-80638":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":15,"forks30d":15,"starsTrendScore":19,"compositeScore":20,"rankGlobal":10,"rankLanguage":10,"license":21,"archived":22,"fork":22,"defaultBranch":23,"hasWiki":22,"hasPages":24,"topics":25,"createdAt":10,"pushedAt":10,"updatedAt":46,"readmeContent":47,"aiSummary":48,"trendingCount":15,"starSnapshotCount":15,"syncStatus":13,"lastSyncTime":49,"discoverSource":50},80638,"agenttrace","luoyuctl\u002Fagenttrace","luoyuctl","Local-first TUI for AI coding-agent session history: trace cost, tokens, time, tool failures, latency, health, diffs, reports, and CI gates across local agent logs.","https:\u002F\u002Fluoyuctl.github.io\u002Fagenttrace\u002F",null,"Go",57,2,7,0,3,4,6,9,50.03,"MIT License",false,"master",true,[26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45],"agent-session-replay","ai-agent-observability","ai-agents","ai-coding","ai-cost-tracking","aider","claude-code","claude-code-logs","cli","cline","codex-cli","cursor","developer-tools","gemini-cli","kimi-cli","opencode","qwen-code","terminal","token-usage","tui","2026-06-12 04:01:29","\u003Cp align=\"center\">\n  \u003Cimg src=\"assets\u002Flogo-icon.png\" alt=\"agenttrace logo\" width=\"256\" height=\"256\">\n\u003C\u002Fp>\n\n\u003Ch1 align=\"center\">AgentTrace\u003C\u002Fh1>\n\n\u003Cp align=\"center\">\n  Local-first TUI and reports for AI coding-agent session history, cost, tokens, time, and slow-run diagnosis.\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  English | \u003Ca href=\"README.zh-CN.md\">简体中文\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fluoyuctl\u002Fagenttrace\u002Factions\u002Fworkflows\u002Fci.yml\">\u003Cimg src=\"https:\u002F\u002Fgithub.com\u002Fluoyuctl\u002Fagenttrace\u002Factions\u002Fworkflows\u002Fci.yml\u002Fbadge.svg\" alt=\"CI\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fluoyuctl.github.io\u002Fagenttrace\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fsite-agenttrace-54ff00.svg\" alt=\"Site\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fluoyuctl\u002Fagenttrace\u002Freleases\u002Flatest\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fv\u002Frelease\u002Fluoyuctl\u002Fagenttrace?color=00ADD8\" alt=\"Release\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fpkg.go.dev\u002Fgithub.com\u002Fluoyuctl\u002Fagenttrace\">\u003Cimg src=\"https:\u002F\u002Fpkg.go.dev\u002Fbadge\u002Fgithub.com\u002Fluoyuctl\u002Fagenttrace.svg\" alt=\"Go Reference\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgoreportcard.com\u002Freport\u002Fgithub.com\u002Fluoyuctl\u002Fagenttrace\">\u003Cimg src=\"https:\u002F\u002Fgoreportcard.com\u002Fbadge\u002Fgithub.com\u002Fluoyuctl\u002Fagenttrace\" alt=\"Go Report Card\">\u003C\u002Fa>\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fgo-1.25+-00ADD8.svg\" alt=\"Go\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-MIT-green.svg\" alt=\"License\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FHomebrew-v0.5.4-2bbc8a.svg\" alt=\"Homebrew\">\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"assets\u002Freadme-real-run.gif\" alt=\"agenttrace running locally against real AI coding agent session logs\" width=\"100%\">\n\u003C\u002Fp>\n\n---\n\n**agenttrace** is a local-first terminal TUI and report generator for AI coding-agent session history. It reads Claude Code, Codex CLI, Gemini CLI, Qwen Code, Cline, Aider, Cursor exports, Hermes Agent, OpenCode, OpenClaw, Pi, Oh My Pi, Kimi CLI, Copilot-style logs, and generic JSON\u002FJSONL traces, then helps with two daily jobs: see what multiple agents spent across cost, tokens, and time; and diagnose why a task ran slowly.\n\n## Why agenttrace?\n\nAI coding agents now behave like small build systems: they call tools, retry, stall, and spend tokens while you only see the final answer.\n\n**agenttrace** reads the logs your agents already write and puts cost-heavy or slow sessions first.\n\nIt helps you answer:\n\n- **What did my agents spend?** Compare historical sessions by agent source, model, input\u002Foutput\u002Fcache tokens, estimated cost, and wall-clock time.\n- **Why was this task slow?** Catch long gaps, hanging sessions, retry loops, slow tool calls, large parameters, and context pressure.\n- **Did a run regress?** Compare against a local baseline when supplied, then inspect incident timelines and conservative tool authority categories in reports.\n- **What should I inspect first?** Rank sessions by cost, duration, turns, health, failures, anomalies, model, source, or text search.\n- **Can I inspect this privately?** Everything runs locally; prompts, code, and logs do not need to leave your machine.\n\n## Real local run\n\nThese screenshots were captured from a local run against real session logs. They are not `--demo` output and not test fixtures.\n\n```bash\nagenttrace\n```\n\n| Overview | Critical sessions |\n|---|---|\n| \u003Cimg src=\"assets\u002Freadme-real-overview.png\" alt=\"agenttrace overview showing real local AI coding agent sessions, token cost, errors, and health\" width=\"100%\"> | \u003Cimg src=\"assets\u002Freadme-real-critical.png\" alt=\"agenttrace critical session list from real local AI coding agent logs\" width=\"100%\"> |\n\n| Session detail | Diagnostics |\n|---|---|\n| \u003Cimg src=\"assets\u002Freadme-real-detail.png\" alt=\"agenttrace detail view showing health, cost, tool failures, and next action from a real local session\" width=\"100%\"> | \u003Cimg src=\"assets\u002Freadme-real-diagnostics.png\" alt=\"agenttrace diagnostics view showing latency, context window, and large parameter calls from real local logs\" width=\"100%\"> |\n\nThat local run found:\n\n```text\nAGENTTRACE v0.5.4\n```\n\n| Signal | What agenttrace found |\n|---|---:|\n| Analyzed sessions | 1,761 |\n| Total tokens | 9.13B |\n| Estimated cost | $5,037.26 |\n| Tool failure rate | 1.1% |\n| Critical sessions | 16 |\n| Average health | 91% |\n\n## Install\n\n```bash\ncurl -sL https:\u002F\u002Fraw.githubusercontent.com\u002Fluoyuctl\u002Fagenttrace\u002Fmaster\u002Finstall.sh | sh\n```\n\nOther install paths:\n\n```bash\nbrew install luoyuctl\u002Ftap\u002Fagenttrace\ngo install github.com\u002Fluoyuctl\u002Fagenttrace\u002Fcmd\u002Fagenttrace@latest\n```\n\nWindows:\n\n```powershell\niwr -useb https:\u002F\u002Fraw.githubusercontent.com\u002Fluoyuctl\u002Fagenttrace\u002Fmaster\u002Finstall.ps1 | iex\n```\n\n## Common workflows\n\n```bash\n# Open the local TUI\nagenttrace\n\n# Check detected agent directories and cache state\nagenttrace --doctor\n\n# Generate machine-readable evidence\nagenttrace --overview -f json\n\n# Search local session metadata without indexing prompt text\nagenttrace --search billing\nagenttrace --search internal\u002Fws -f json\n\n# Create a self-contained report for CI artifacts or issue links\nagenttrace --overview -f html -o agenttrace-overview.html\n\n# Save a local baseline, then compare a later run\nagenttrace --overview -f json -o agenttrace-baseline.json\nagenttrace --overview -f json \\\n  --baseline agenttrace-baseline.json \\\n  -o agenttrace-overview.json\n\n# Fail CI on unhealthy agent runs\nagenttrace --overview \\\n  --fail-under-health 80 \\\n  --fail-on-critical \\\n  --max-tool-fail-rate 15\n```\n\n## Supported logs\n\nagenttrace supports local sessions from:\n\nClaude Code, Codex CLI, Gemini CLI, Qwen Code, Cline, Aider, Cursor exports, Hermes Agent, OpenCode, OpenClaw, Pi, Oh My Pi, Kimi CLI, Copilot-style logs, and generic JSON\u002FJSONL traces.\n\n## What you get\n\n| Need | agenttrace gives you |\n|---|---|\n| Historical spend review | Sessions grouped across agents with token totals, model pricing, estimated cost, and elapsed time |\n| Slow-task diagnosis | Latency stats, long gaps, hanging sessions, retry loops, slow tools, large params, and context pressure |\n| Regression evidence | Local baseline comparison when supplied, incident timelines, and conservative tool authority categories in reports |\n| First-session triage | Sort and filter by cost, duration, health, failures, anomalies, model, source, or text search |\n| Shareable evidence | JSON, Markdown, and self-contained HTML reports |\n| Local-first inspection | No hosted backend required |\n\n## Docs\n\n- Site: https:\u002F\u002Fluoyuctl.github.io\u002Fagenttrace\u002F\n- Sample HTML report: https:\u002F\u002Fluoyuctl.github.io\u002Fagenttrace\u002Fdemo-report.html\n- CI setup: [docs\u002Fci-integration.md](docs\u002Fci-integration.md)\n- Cursor import: [docs\u002Fcursor-import.md](docs\u002Fcursor-import.md)\n- Parser guide: [docs\u002Fparser-guide.md](docs\u002Fparser-guide.md)\n- Launch notes: [docs\u002Flaunch-kit.md](docs\u002Flaunch-kit.md)\n\nListed in:\n\n- [Awesome Gemini CLI](https:\u002F\u002Fgithub.com\u002FPiebald-AI\u002Fawesome-gemini-cli)\n- [Charm in the Wild](https:\u002F\u002Fgithub.com\u002Fcharm-and-friends\u002Fcharm-in-the-wild)\n- [Awesome Claude Code and Skills](https:\u002F\u002Fgithub.com\u002FGetBindu\u002Fawesome-claude-code-and-skills)\n- [awesome-x-ops](https:\u002F\u002Fgithub.com\u002Fxlabs-club\u002Fawesome-x-ops)\n- [Awesome DevOps AI](https:\u002F\u002Fgithub.com\u002Fhammadhaqqani\u002Fawesome-devops-ai)\n- [agentic-ai-knowledge-base](https:\u002F\u002Fgithub.com\u002Fankurkumarz\u002Fagentic-ai-knowledge-base)\n- [awesome-harness-engineering](https:\u002F\u002Fgithub.com\u002Fwalkinglabs\u002Fawesome-harness-engineering)\n- [awesome-ai-tools](https:\u002F\u002Fgithub.com\u002FQAInsights\u002Fawesome-ai-tools)\n- [awesome-claude-code-toolkit](https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fawesome-claude-code-toolkit)\n- [awesome-ai-plugins](https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-ai-plugins)\n- [awesome-agent-skills](https:\u002F\u002Fgithub.com\u002Fkodustech\u002Fawesome-agent-skills)\n\n## Contributing\n\nParser PRs are welcome. A good parser contribution usually includes:\n\n- a tiny redacted fixture or synthetic sample\n- format detection in `DetectFormat`\n- role, timestamp, model, token usage, tool call, and tool error extraction\n- tests for successful parsing and malformed input\n\nRun before sending a PR:\n\n```bash\ngo test .\u002F...\ngo build -o agenttrace .\u002Fcmd\u002Fagenttrace\u002F\n.\u002Fagenttrace --doctor\n```\n\nSee [CONTRIBUTING.md](CONTRIBUTING.md) for the full contribution flow.\n\n## License\n\n[MIT](LICENSE) © 2026 agenttrace contributors\n","agenttrace 是一个面向本地的终端TUI和报告生成器，用于AI编码代理会话历史记录。它能够读取包括Claude Code、Codex CLI、Gemini CLI、Qwen Code等在内的多种AI编码代理的日志，并帮助用户追踪成本、令牌使用量以及时间消耗情况，同时诊断运行缓慢的原因。其核心技术特点在于支持广泛的日志格式输入及提供详细的性能分析报告。适用于需要对AI辅助编程工具进行成本控制与性能优化的开发者或团队，在保证数据隐私的同时提高开发效率。","2026-06-11 04:01:28","CREATED_QUERY"]