# seatledger (npm · seatledger)

Local ledger of Claude Code and Codex tokens, API-equivalent cost, limits and token-rate changes.

- Trust score: 62/100 (medium)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-10-17

## Components

- npm · `seatledger`: 62/100 (this document), [markdown](https://verifymcp.io/servers/agentwares-seatledger/seatledger.md), [page](https://verifymcp.io/servers/agentwares-seatledger/seatledger)

## Channel facts

- Registry: `npm`
- Package: `seatledger`
- Version: `0.3.0`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-10-17.

- **Supply Chain Security**: 100/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - No install/post-install scripts declared.
  - No production dependencies, so there is no dependency health to assess.
- **Provenance & Transparency**: 19/100
  - Repository check failed: the declared repository URL returned HTTP 404.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 8 days ago).
  - Security-disclosure policy not yet verified: we couldn't inspect the source repository.
- **Schema Quality & AI Usability**: 59/100
  - AI-judged instruction clarity (excellent).
  - Context-footprint check failed: tool/resource definitions use about 1351 tokens (~337/item across 4 items; 4 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 0/100
  - Stability not yet verified: not enough scan history yet (needs a 30-day window).
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 4 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 5 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

**Unverified: 1 category.** A category scored 0 because we could not verify it: a data source with nothing on this package, evidence we could not reach, or a check we could not run. We only credit what we can confirm.

## Install

### How do I install the seatledger MCP server?

seatledger runs locally as an npm package, launched with npx -y seatledger. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add agentwares-seatledger -- npx -y seatledger
```

### Cursor

```json
{
  "mcpServers": {
    "agentwares-seatledger": {
      "command": "npx",
      "args": [
        "-y",
        "seatledger"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "agentwares-seatledger": {
      "command": "npx",
      "args": [
        "-y",
        "seatledger"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add agentwares-seatledger -- npx -y seatledger
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "agentwares-seatledger": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "seatledger"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add agentwares-seatledger --command npx --arg -y --arg seatledger
```

### Hermes

```yaml
mcp_servers:
  agentwares-seatledger:
    command: "npx"
    args: ["-y", "seatledger"]
```

### Netclaw

```json
{
  "McpServers": {
    "agentwares-seatledger": {
      "Transport": "stdio",
      "Command": "npx",
      "Arguments": [
        "-y",
        "seatledger"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add agentwares-seatledger -t stdio -c npx -a -y seatledger
```

### Other

```json
{
  "mcpServers": {
    "agentwares-seatledger": {
      "command": "npx",
      "args": [
        "-y",
        "seatledger"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-10-14 (score 62, +28)

- [security improvement] Injection markers: unverified → pass
- [security] First check of Judged manipulation: pass
- [security] Stability: Stability not yet verified: not enough scan history yet (needs a 30-day window).
- [functional improvement] Tool coverage: unverified → 100
- [functional improvement] MCP protocol: unverified → pass
- [functional] First check of Schema quality: fail
- [functional] First check of Schema quality: excellent
- [functional] First check of Destructive annotations: pass
- [functional] First check of Schema quality: fail
- [functional] First check of Tool coverage: 100

### 2026-10-11 (score 34)

First indexed and scored.

## MCP tools (4)

### `seatledger_usage_summary` (~301 tokens)

Coding-agent usage on this machine

Tokens and API-equivalent dollars from this machine's Claude Code and Codex transcripts, read locally. With no arguments: today by client and model, the last 7 days by day, progress against limits the user recorded with `seatledger limits set`, and the latest rate-limit readings Codex wrote to its own rollouts. With `since` or `group_by`: one table for that range, grouped by day, project folder, model or client. Tokens count input, output, cache reads and cache writes. API-equivalent means the same tokens at the vendor's API list prices (dated table, sources included); it is not what the user was billed, since a seat is a flat fee with an allowance. Requests on models without a list price are counted in tokens and listed as unpriced.

Input parameters:

- `client` (string): Only this coding-agent client: `claude-code` (Claude Code transcripts) or `codex` (Codex CLI rollouts). Omitted: both.
- `group_by` (string): Group the range by local `day`, `project` (working-directory folder name), `model` or `client`. Omitted with `since` set: day.
- `since` (string): Start of the range: a local date (`2026-09-01`) or a span back from today (`30d`, `12w`). Omitted with `group_by` set: 30d.

### `seatledger_rate_changes` (~284 tokens)

Token rate changes

Step changes in how this machine's coding-agent requests use tokens, per client and model: tokens per request, cache-read share of prompt tokens, cache writes per cache read, tokens per user turn, the share of Claude Code cache writes made at the 1-hour lifetime, and Codex tokens per 1% of its 5-hour window. Each finding has the date it starts, the client version in use and whether that was its first day, before and after medians, and what the change is consistent with. It sees only this machine's requests, not the vendor's servers, so it never states a cause. Also returns each model's current baseline and the method. Days with fewer than 20 requests are left out.

Input parameters:

- `client` (string): Only this coding-agent client: `claude-code` (Claude Code transcripts) or `codex` (Codex CLI rollouts). Omitted: both.
- `model` (string): Only this model id, exactly as the client records it (e.g. `claude-opus-5-5`, `gpt-5.6-sol`). Omitted: every model.
- `since` (string): Start of the history to examine: a local date (`2026-07-01`) or a span back from today (`90d`). Omitted: 90d.

### `seatledger_plan_compare` (~402 tokens)

This machine's month on every coding plan

What this machine's last 30 days of Claude Code and Codex work would cost on each coding plan, read from the local transcripts: the Anthropic and OpenAI APIs at list prices; Claude Pro, Max 5x, Max 20x, Team standard and Team premium; ChatGPT Plus and Pro; Codex credits at OpenAI's rate card; and Copilot Pro, Pro+ and Max at GitHub's per-token AI-credit rates. Returns each client's tokens, API-equivalent dollars and busiest 5 hours and week; a quote per plan saying how it was reached (priced from published rates, measured from Codex's own window readings, or an estimate); per client, the cheapest priced plan, the runner-up and the margin; plans that could be cheaper if an allowance the vendor does not publish holds the busiest 5 hours and week; and the newest dated vendor change that moved the answer. Allowances Anthropic and OpenAI do not publish as numbers are never invented: those plans get their fee and that condition. Copilot figures are estimates, since Copilot keeps no local logs and its agent would not use exactly these tokens. The dollar price of a Codex credit is derived from OpenAI's own figures and marked so. Every plan carries its source URLs and the date they were read. Costs are stated for a 30-day month.

Input parameters:

- `plans` (array): Only these plans: plan ids such as `claude-max-5x`, `chatgpt-plus` or `copilot-max`, or a vendor group (`claude`, `codex`, `copilot`). Omitted: every plan.
- `since` (string): Start of the range: a local date (`2026-09-01`) or a span back from today (`30d`, `12w`). Metered figures are scaled to a 30-day month either way. Omitted: the last 30 days.

### `seatledger_team_summary` (~275 tokens)

Team coding-agent ledger

The hosted seatledger team ledger this machine created or joined (`seatledger team create` or `seatledger team join`): what the team's coding-agent seats and tokens bought, from every developer's pushed daily aggregates. Returns team totals; tables by day, month, developer, model and client; and the rate-change findings feed, each with its date, the developer, the client version and whether that was its first day, and what the step is consistent with. A developer's key sees every team total and only their own row in `by_developer`; the owner key sees every developer. Dollars are API-equivalent (the same tokens at API list prices), not what anyone was billed. History is as long as the team's plan keeps (Free 30 days, Team and Business 13 months). Needs the key `seatledger team create` or `team join` saved in ~/.seatledger/team.json, or SEATLEDGER_KEY; without one it returns NO_TEAM_KEY. This tool makes one HTTPS request to that ledger and changes nothing.

Input parameters:

- `since` (string): Start of the range: a date (`2026-09-01`) or a span back from today (`30d`, `12w`, `13m`). Never earlier than the plan keeps. Omitted: everything the plan keeps.

## Diagnostics

Captured diagnostic sections: Provenance, Dependencies. The full working is on the page: https://verifymcp.io/servers/agentwares-seatledger/seatledger#diagnostics

## Score history

- 2026-10-17: 62
- 2026-10-16: 62
- 2026-10-15: 62
- 2026-10-14: 62
- 2026-10-13: 34
- 2026-10-12: 34
- 2026-10-11: 34

## Common questions

### What is the seatledger MCP server?

seatledger is an MCP server listed in the public MCP registry as io.github.agentwares/seatledger. Local ledger of Claude Code and Codex tokens, API-equivalent cost, limits and token-rate changes. This page covers its npm package (seatledger).

### Is the seatledger MCP server safe to use?

seatledger scores 62 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 17 October 2026. It declares no install or post-install scripts. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the seatledger MCP server expose?

seatledger exposes 4 tools: seatledger_usage_summary, seatledger_rate_changes, seatledger_plan_compare, seatledger_team_summary. Their descriptions and schemas cost roughly 1,262 tokens of context every time the server is loaded.

### Is the seatledger MCP server still maintained?

seatledger is still listed as active in the MCP registry. We last reached this channel on 17 October 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What licence is the seatledger MCP server under?

seatledger declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

## Links

- npm package: https://www.npmjs.com/package/seatledger
- Socket report: https://socket.dev/npm/package/seatledger
- Changelog RSS feed: https://verifymcp.io/servers/agentwares-seatledger/seatledger.xml
- Changelog JSON feed: https://verifymcp.io/servers/agentwares-seatledger/seatledger.json
- HTML version of this page: https://verifymcp.io/servers/agentwares-seatledger/seatledger