ai-usage-mcp

MCPattested

v0.5.1Mohit BansalMITUpdated 2d agonpm GitHub

Local-first MCP server that reports real token usage and cost across coding agents (Claude Code, OpenCode). Nothing leaves your machine.

Works in

ClaudeCursorCopilotGemini

Inferred from the transports this listing declares (stdio). A client not listed here hasn’t been ruled out — it just isn’t something Forge can confirm.

Attested build

A verified provenance attestation binds this artifact to the listed repository. Nobody has claimed the listing yet — this proves where the code was built, not who stands behind it.

1GitHub stars

2d agoLast update

Package

AuthorMohit Bansal

LicenseMIT

Version0.5.1

Sourcenpm+mcp-registry

npmai-usage-mcp

GitHubMohitBansal321/ai-usage-mcp

SBOMCycloneDX 1.6 ↓

Trust Status

A

85/100Trusted

✓Listed in Forge index+10/10

✓Identity verified · attested build+20/20

—Ed25519 publish signature+0/5

→ Included automatically when the publisher runs `forge publish`

—Domain verification+0/5

→ Publisher: host /.well-known/forge.json on the package homepage with { "publisher": "<github-login>" }

✓npm Trusted Publishing (Sigstore)+5/5

—npm maintainer match+0/5

→ Publisher: add the verified GitHub login to the npm package's maintainers (npm owner add <login>)

✓CVE scan · clean+30/30

✓Static analysis · clean+20/20

Paste into Claude Code, Cursor, or any AI assistant to fix all gaps

StatusIdentity verified

PublisherUnverified

SignatureUnsigned

Domain—

Provenance✓ Sigstore-verified · 58ef299

Dependencies✓ 60 resolved+ · none vulnerable

Tool surface7 tools · none privileged

Security scan✓ Cleanv0.5.1 · todayHow well does this scan work?

EvalsNone

IndexedSep 5, 2026

Verification confirms publisher identity (repo ownership), not code safety. The security scan covers known CVEs and suspicious install scripts.

Tools

7 tools · none privileged

Statically extracted from the published packagev0.5.1 · 2h ago

Read out of the source npm actually ships, at scan time. The package was never executed. Tools registered dynamically at runtime, or hidden inside bundled or minified code, can be missed — so this is a floor on the tool surface, not a complete census of it.

Filter toolsPrivileged only

client_usageToken usage and cost per coding agent (Claude Code vs OpenCode).

Token usage and cost per coding agent (Claude Code vs OpenCode).

No input schema was published for this tool.

daily_usagePer-day token usage and cost, newest day first. Use this for "how much did I use

Per-day token usage and cost, newest day first. Use this for "how much did I use

No input schema was published for this tool.

model_usagePer-model token usage and cost, highest token count first. Use this to answer

Per-model token usage and cost, highest token count first. Use this to answer

No input schema was published for this tool.

project_usagePer-project token usage and cost, highest token count first. A project is the working

Per-project token usage and cost, highest token count first. A project is the working

No input schema was published for this tool.

recent_sessionsMost recently active sessions with project path, client, model(s), duration,

Most recently active sessions with project path, client, model(s), duration,

No input schema was published for this tool.

session_usageUsage for one session: client, model(s), duration, token breakdown and cost.

Usage for one session: client, model(s), duration, token breakdown and cost.

No input schema was published for this tool.

usage_summaryTotal token usage and cost for a period, split by client (Claude Code, OpenCode).

Total token usage and cost for a period, split by client (Claude Code, OpenCode).

No input schema was published for this tool.

7 of 7 tools published a description.

Tool names and descriptions are written by the publisher and shown verbatim as inert text. They are the strings an MCP client passes to a model, so Forge scans them for prompt-injection patterns — any finding appears with the security scan above. “Privileged” is a keyword match on the tool name, not an audit of what the tool does: a benign-sounding name can still do anything.

About

Local-first MCP server that reports real token usage and cost across coding agents (Claude Code, OpenCode). Nothing leaves your machine.

Keywords

mcpmodel-context-protocolclaude-codeopencodetoken-usagecostlocal-first

Alternatives

Comparing tool surfaces…

Tools exposed

7 tools · none privileged · observed 2026-09-05 by static extraction. A tool surface is what one scan read on one day, not a live capability list.

client_usage
daily_usage
model_usage
project_usage
recent_sessions
session_usage
usage_summary

Linked names open Forge’s index of every entry observed exposing that tool. Browse all indexed tools.

Dependency tree

What one Forge scan resolved from npm metadata on 2026-09-05 — observed resolution, not a publisher declaration.

60 packages resolved · 2 direct · none carrying advisories Resolution stops at depth 4 and 60 packages.

zod4.5.4
@modelcontextprotocol/sdk1.30.017 dependencies
- ajv8.20.04 dependencies
  - fast-deep-equal3.1.3
  - fast-uri3.1.7
  - json-schema-traverse1.0.0
  - require-from-string2.0.2
- zod4.5.4shown above
- cors2.8.62 dependencies
  - object-assign4.1.1
  - vary1.1.2
- hono4.13.7
- jose6.2.12
- express5.2.1+26 not drawn2 dependencies
  - vary1.1.2shown above
  - content-type1.0.5
- raw-body3.0.2+4 not drawn
- ajv-formats3.0.11 dependency
  - ajv8.20.0shown above
- cross-spawn7.0.6+3 not drawn
- eventsource3.0.71 dependency
  - eventsource-parser3.1.1
- content-type1.0.5shown above
- pkce-challenge5.0.1
- @hono/node-server2.1.1
- json-schema-typed8.0.2
- eventsource-parser3.1.1shown above
- express-rate-limit8.7.0+2 not drawn
- zod-to-json-schema3.25.2

The crawl stopped at the depth-4 limit. Anything below that level was never resolved.

The crawl stopped at the 60-package limit. The rest of the tree was never resolved.

36 more resolved packages are not drawn here (display cap: 24). Every dependency carrying an advisory is drawn regardless of the cap. Full inventory (CycloneDX SBOM)

Declared but not resolved

53 declared dependencies never landed in the tree. They are missing from Forge's resolution, not from the package.

package cap reached: 48
below the depth cap: 5

http-errors ^2.0.1 — required by send@1.2.1package cap reached
mime-types ^3.0.2 — required by send@1.2.1package cap reached
ms ^2.1.3 — required by send@1.2.1package cap reached
on-finished ^2.4.1 — required by send@1.2.1package cap reached
range-parser ^1.2.1 — required by send@1.2.1package cap reached
statuses ^2.0.2 — required by send@1.2.1package cap reached
ms ^2.1.3 — required by debug@4.4.3package cap reached
debug ^4.4.0 — required by router@2.2.0package cap reached
depd ^2.0.0 — required by router@2.2.0package cap reached
is-promise ^4.0.0 — required by router@2.2.0package cap reached
parseurl ^1.3.3 — required by router@2.2.0package cap reached
path-to-regexp ^8.0.0 — required by router@2.2.0package cap reached

+41 more not listed. The counts by reason above cover all of them.

Not followed: peerDependencies. This tree covers runtime dependencies only, so anything those pull in was never resolved.