agenttrace

Name: agenttrace
Author: luoyuctl

Verified

Local-first Rust TUI/CLI for auditing AI coding-agent sessions: cost, tokens, latency, failures, and health｜本地优先 Rust TUI/CLI，审计 AI 编程 Agent 会话的成本、Token、延迟、失败与健康度

119stars

6forks

Rust

Installation

# Add to your Claude Code skills
git clone https://github.com/luoyuctl/agenttrace

Getting Started

Guides for using ai agents skills like agenttrace.

Caveman: Cut Claude Token Use by 65%
How agent-side prompt compression works, when to use it, and when not to.
What is an AI Skills Marketplace?
Definitions, how marketplaces work, and how to choose between them in 2026.
Getting Started with AI Skills

Security ReportVerified

Last scanned: 7/14/2026

{
  "issues": [
    {
      "file": "README.md",
      "line": 82,
      "type": "remote-install",
      "message": "Install command (remote install script piped to a shell — review the source before running): \"curl -sL https://raw.githubusercontent.com/luoyuctl/agenttrace/master/install.sh\"",
      "severity": "low"
    }
  ],
  "status": "PASSED",
  "scannedAt": "2026-07-14T06:10:39.698Z",
  "npmAuditRan": true,
  "pipAuditRan": true,
  "promptInjectionRan": true
}

README.md

Frequently Asked Questions

What is agenttrace?

agenttrace is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by luoyuctl. Local-first Rust TUI/CLI for auditing AI coding-agent sessions: cost, tokens, latency, failures, and health｜本地优先 Rust TUI/CLI，审计 AI 编程 Agent 会话的成本、Token、延迟、失败与健康度. It has 119 GitHub stars.

Is agenttrace safe to use?

Yes. agenttrace passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.

How do I install agenttrace?

Clone the repository with "git clone https://github.com/luoyuctl/agenttrace" and add it to your Claude Code skills directory (see the Installation section above).

What programming language is agenttrace written in?

agenttrace is primarily written in Rust. It is open-source under luoyuctl on GitHub, so you can review or fork the full source.

Are there alternatives to agenttrace?

Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh agenttrace against similar tools.

Agentic AI for Beginners

Build your first AI agent from scratch - tool use, ReAct pattern, memory, deployment

41 minBeginner

Comments (0)

to leave a comment.

No comments yet. Be the first to share your thoughts!

Related Skills

ECC

by affaan-m

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

240,740

Developers Also Liked

Based on votes and bookmarks from developers who liked this skill

hermes-agent

by NousResearch

The agent that grows with you

232,110

agent-watch-approve event-tracking-skill

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/JSONL traces, then helps with two daily jobs: see what multiple agents spent across cost, tokens, and time; and diagnose why a task ran slowly.

One agenttrace binary provides both interfaces: run it without a report action to open the TUI, or pass flags such as --sessions and --overview for CLI output.

Why agenttrace?

AI coding agents now behave like small build systems: they call tools, retry, stall, and spend tokens while you only see the final answer.

agenttrace reads the logs your agents already write and puts cost-heavy or slow sessions first.

It helps you answer:

What did my agents spend? Compare historical sessions by agent source, model, input/output/cache tokens, estimated cost, and wall-clock time.
Why was this task slow? Catch long gaps, hanging sessions, retry loops, slow tool calls, large parameters, and context pressure.
Did a run regress? Compare against a local baseline when supplied, then inspect incident timelines and conservative tool authority categories in reports.
What should I inspect first? Rank sessions by cost, duration, turns, health, failures, anomalies, model, source, or text search.
Can I inspect this privately? Everything runs locally; prompts, code, and logs do not need to leave your machine.

Real local run

These screenshots and figures are a redacted sample of the latest 500 real local sessions captured for the v0.7.1 source tree. They are not --demo output, current telemetry, or test fixtures.

agenttrace

Overview	Critical sessions

Session detail	Diagnostics

That local run found:

AGENTTRACE v0.7.1

Signal	What agenttrace found
Analyzed sessions	500
Total tokens	4.0B
Estimated cost	$2.6K
Tool failure rate	2.9%
Critical sessions	2
Average health	77.7%

Install

Install the latest public release with your package manager. Check the installed version with agenttrace --version.

# macOS and Linux
brew install luoyuctl/tap/agenttrace

# macOS, Linux, and Windows (requires Node.js 18+)
npm install -g @zack78/agenttrace

Windows:

winget install --id Luoyuctl.AgentTrace --exact

The package names above become available once the corresponding release has been published. Manual installs remain available without a package manager:

curl -fsSL https://raw.githubusercontent.com/luoyuctl/agenttrace/master/install.sh | sh
cargo install --git https://github.com/luoyuctl/agenttrace agenttrace

iwr -useb https://raw.githubusercontent.com/luoyuctl/agenttrace/master/install.ps1 | iex

Quickstart

agenttrace

Governance reports

# Audit raw token components, normalized pricing, and fallback confidence.
agenttrace --audit --range 30d -f json

# Optional local model aliases and per-million-token price overrides.
AGENTTRACE_PRICING_FILE=pricing-overrides.json agenttrace --audit -f json

# Rank evidence-backed actions by severity and estimated impact.
agenttrace --recommend --range 30d -f json

# Inspect observed MCP invocations. Loaded-server coverage is intentionally
# reported as unavailable unless the source log actually records it.
agenttrace --mcp-governance --range 30d -f json

# Review cross-session context, cache, repeat-read, and read/write trends.
agenttrace --context-trends --range 30d -f json

# Correlate local Git commit timestamps with sessions (heuristic, read-only).
agenttrace --delivery-evidence --range 30d -f json

pricing-overrides.json accepts aliases plus per-million-token prices:

{"aliases":{"provider/raw-model":"my-model"},"prices":{"my-model":{"input":1,"output":2,"cw":0,"cr":0}}}

--overview now includes scope, parse and pricing confidence, cost audit, prioritized recommendations, MCP governance, context trends, and delivery signals in JSON, Markdown, and HTML output. All cost and delivery fields are explicitly estimates or heuristics; they are not provider billing or proof that a commit reached main.

What you get

Need	agenttrace gives you
Historical spend review	Sessions grouped across projects, agents, and models with Today/7d/30d/All ranges
Data confidence	Report scope, per-source coverage, parse skips, cache hits, unknown sources/models, pricing fallbacks, and latest observed session
Cost audit and action plan	Token component rates, pricing source/status, estimated cost confidence, and prioritized, evidence-backed remediation suggestions
Governance trends	Canonical project grouping, observed MCP invocation governance, cross-session context/cache/read-write trends, and read-only Git delivery correlation
Honest capability levels	`Detailed`, `Aggregate`, or `Limited` per session so missing event-level evidence is never presented as a complete trace
Privacy-safe steps	Tool-step metadata and duration when the source provides call IDs and timestamps; no prompt, response, result, or tool-argument body is stored in steps
Slow-task diagnosis	Latency stats, long gaps, hanging sessions, retry loops, slow tools, large params, and context pressure
Regression evidence	Local baseline comparison when supplied, incident timelines, and conservative tool authority categories in reports
First-session triage	Sort and filter by cost, duration, health, failures, anomalies, model, source, or text search
Shareable evidence	JSON, Markdown, and self-contained HTML reports
Local-first inspection	No hosted backend required

Docs

Documentation index: docs/README.md
CI setup: docs/guides/ci-integration.md
Governance reports: docs/guides/governance-reports.md
Cursor import: docs/guides/cursor-import.md
Parser guide: docs/guides/parser-guide.md
Maintainer distribution guide: docs/maintainers/distribution.md

Listed in these open source projects:

Contributing

Parser PRs are welcome. A good parser contribution usually includes:

a tiny redacted fixture or synthetic sample
format detection in crates/agenttrace-core/src/parser.rs
role, timestamp, model, token usage, tool call, and tool error extraction
tests for successful parsing and malformed input

Run before sending a PR:

cargo test
cargo build --release -p agenttrace
target/release/agenttrace --doctor

See CONTRIBUTING.md for the full contribution flow.