token-meter
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| prompts | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| usage_summaryA | What you spent, where it went, and what was slow — Claude Code + Codex. API-equivalent estimate, local data only. Table by day × models (ccusage-style) + advisory narrative. scope defaults to "auto" (filter to current platform); pass "all" to see everything. insights=true adds heuristic tips. |
| recent_sessionsB | List recently-active Claude Code / Codex sessions with ready-to-paste resume commands. |
| session_toolsA | Per-session MCP / built-in tool breakdown: call counts, response sizes, average latency. Debug a slow/expensive session. |
| subagent_costsA | Split spend into main-session vs sub-agent (Task/Agent) work, list the priciest sub-agents (model mix · tokens · cache), and pair them with parent-side invocation latency. Answers "are my sub-agents worth what they cost". Run |
| refresh_dataA | Re-scan local Claude Code / Codex JSONL for new activity. Run before other tools for up-to-the-minute numbers. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| usage_summary | Summarize Claude Code + Codex usage for today / week / month. |
| recent_sessions | List recent Claude Code / Codex sessions with ready-to-paste resume commands. |
| session_tools | Show MCP / built-in tool breakdown for a specific session_id. |
| subagent_costs | Show main vs sub-agent spend split and the priciest sub-agents. |
| refresh_data | Re-scan local JSONL files to pick up new Claude Code / Codex activity. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
Each tool has a distinct focus: aggregate usage, session listing, per-session tool stats, subagent cost split, and data refresh. There is mild overlap between usage_summary and subagent_costs since both report spend, but the descriptions clearly delineate the scope.
All names use snake_case and mostly follow a noun-phrase pattern like usage_summary, recent_sessions, and session_tools. refresh_data breaks the pattern as an imperative verb, but the overall naming style is still coherent and predictable.
Five tools is a well-scoped count for a focused usage/cost monitoring server. Each tool addresses a distinct analytical need without unnecessary bloat.
The tool surface covers the core workflow: refresh local data, view aggregate usage, list sessions, inspect per-session tool activity, and analyze subagent costs. Missing advanced capabilities like time-range filtering or raw data export feel like minor gaps rather than fundamental omissions.