Mmcp.market

TokenTrace MCP server

by abhiyoheswaran1·io.github.abhiyoheswaran1/tokentrace·v0.21.2
No reviews yet

Local-first AI CLI usage analytics for agents.

B82/100grade B
Write a reviewInstall
What users say
No reviews yet
Be the first
Safety scan
B82/100

full report

Adoption
New

0 stars62 downloads/wk

Reviews

Write one

Nobody has reviewed TokenTrace yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

TokenTrace tools (13, 3 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • get_agent_guide

    Return the recommended MCP operating loop, install snippets, copy-paste AGENTS.md guidance, and local-first guardrails.

  • get_anomalies

    Return modified-z-score (MAD) anomalies for the local daily token and cost trend. Pure-stats detector; spends zero AI tokens.

  • get_capabilities

    Return the read-only TokenTrace agent discovery manifest. Does not initialize local app data.

  • get_classifications

    Return the local unknown-cost repair queue with deterministic classification suggestions (exact-model, family-fragment, or parser-source). Read-only; spends zero AI tokens.

  • get_evidence

    Return evidence behind a TokenTrace metric without raw prompt or message bodies.

  • get_handoff

    Return the schema-versioned tokentrace.handoff.v1 envelope summarizing local scan state, repair queue, confidence, recent agent actions, and suggested next actions. Pure read; no scan.

  • get_preflight

    Return a read-only local preflight decision before another coding-agent run: proceed, caution, or blocked with evidence and next actions.

  • get_repair_queue

    Return grouped unknown-cost repair causes, next actions, and review state.

  • get_report

    Return a deterministic local report as markdown or JSON.

  • get_status

    Return the current local TokenTrace usage status snapshot as JSON.

  • query_usagewrite action

    Run a deterministic, parameterized query over the local TokenTrace database. The caller supplies structured arguments; TokenTrace executes the SQL. No NL parsing happens server-side, so zero AI tokens are spent.

  • run_doctorwrite action

    Inspect scan freshness, parser trust, source coverage, package trust, and repair recommendations.

  • run_scanwrite action

    Explicitly scan local AI CLI artifacts and update the local TokenTrace database. Requires confirmLocalScan=true.

Public scan report

scanner v0.1.9 · 2026-09-26 · same rubric, same numbers if you re-run it

no findings
  • Code scan307 source files scanned25/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitylocal package, no credentials required12/15
  • Maintenancelast push 92 days ago8/15
  • Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
Overall 82/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

What the publisher says

From the TokenTrace repository's README, as published. We do not edit it. Read it on GitHub

TokenTrace CLI

Local-first AI CLI usage analytics for developers and coding agents.

TokenTrace answers one practical question before the next expensive AI CLI run: is your local usage evidence ready to trust? It scans local Claude Code, Codex, structured usage logs, and usage-shaped local databases, then labels token and cost data as exact, estimated, unknown, cached, or non-cache. The dashboard opens on Today, with direct paths into Sessions, Evidence, and Fix Data.

TokenTrace is designed for local development machines first, with macOS-oriented defaults. It does not require a cloud account and does not send telemetry or logs anywhere.

Website · Source

What TokenTrace Helps You Do

  • Check readiness before another coding-agent run with tokentrace preflight --json or MCP get_preflight.
  • See today’s local cost, tokens, sessions, confidence, anomalies, and repair signals in one first screen.
  • Trace every important number back to local files, parser state, sessions, model rates, and confidence labels.
  • Fix missing cost data through a guided repair queue without uploading prompts or message bodies.

Start In Seconds

Run without installing:

npx tokentrace

Or install globally:

npm install -g tokentrace
tokentrace

The command starts the local dashboard, chooses an available localhost port starting at 3030, opens your default browser, and keeps the server running until you press Ctrl+C.

CLI commands:

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Install directly

Runs npx -y tokentrace on your machine. Read the scan report first; the gateway never runs local packages.

claude mcp add tokentrace -- npx -y tokentrace
Add to Cursor

TokenTrace: common questions

Is TokenTrace MCP server safe?
Mostly: it is graded B (82/100). Read the TokenTrace safety report
How do I install TokenTrace?
It runs on your machine. Copy the Claude Code, Claude Desktop or Cursor config from the install section.
Does TokenTrace need an API key?
Not as far as the registry entry and our scan can tell: no credentials are declared or required.
Is TokenTrace maintained?
The last commit was 93 days ago (2026-06-26). The latest release is v0.21.2.

More from abhiyoheswaran1 →