@sriinnu/drishti (दृष्टि — "vision") is the observation layer for token usage across AI coding agents. It provides an MCP server that exposes token data as tools, a live TUI dashboard, a statusline for editor hooks, and a cross-provider aggregation daemon.

Pairs with @sriinnu/tokmeter for the core parsing engine. Works with Claude Code, Cursor, OpenCode, Codex CLI, Windsurf, Zed, VS Code Copilot, and more.

Install

npm install -g @sriinnu/drishti

Or run directly:

npx @sriinnu/drishti

Commands

| Command | Description | | --------------------------------- | ------------------------------------------- | | drishti | Start live TUI dashboard (default) | | drishti live | Start live TUI dashboard | | drishti serve | Start MCP server (stdio transport) | | drishti statusline | Statusline mode for editor hooks | | drishti daemon start | Start cross-provider aggregation daemon | | drishti daemon stop | Stop the daemon | | drishti daemon status | Check daemon status | | drishti install-statusline | Install statusline hook for all editors | | drishti install-mcp | Install MCP server for all editors | | drishti install-hooks | Install guard hooks (Claude Code) | | drishti install-all | Restore everything — statusline + MCP + hooks | | drishti editors | List all supported editors |

Usage

Live Dashboard

# Launch the real-time TUI observatory
drishti live

Interactive terminal dashboard with live-updating token counts, cost breakdowns, and sparkline trends. Refreshes automatically as sessions change.

MCP Server

# Start as MCP server (stdio transport for editor integration)
drishti serve

MCP Configuration

Add to your editor's MCP settings (e.g., ~/.claude/settings.json):

{
  "mcpServers": {
    "drishti": {
      "command": "npx",
      "args": ["@sriinnu/drishti", "serve"]
    }
  }
}

Or with a global install:

{
  "mcpServers": {
    "drishti": {
      "command": "drishti",
      "args": ["serve"]
    }
  }
}

Auto-install for all supported editors:

drishti install-mcp

MCP Tools

Once connected, drishti exposes these tools to the AI agent:

| Tool | Description | | ------------------- | ---------------------------------------------- | | token_usage | Token usage summary (today / week / month / all-time) | | cost_breakdown | Cost breakdown by model, provider, or project | | daily_trend | Daily usage trend with sparkline | | session_cost | Current session cost and burn rate | | budget_check | Check remaining budget against a limit | | compare_models | Compare cost-efficiency across models | | export_csv | Export usage data as CSV |

Statusline

# Run once for statusline output (designed for editor hooks)
drishti statusline

The statusline produces a compact, ANSI-colored summary of your current session's token usage and cost. Designed to be called by editor hooks (Claude Code, OpenCode, etc.) and rendered inline.

Auto-install for all supported editors:

drishti install-statusline

Daemon

The daemon enables real-time cross-provider aggregation. When running, the statusline shows both your current session totals AND aggregated totals from all open AI coding agents.

# Start the aggregation daemon
drishti daemon start

# Check status
drishti daemon status

# Stop it
drishti daemon stop

Programmatic API

import { startServer } from "@sriinnu/drishti";

// Start MCP server programmatically
await startServer();

import { startLive } from "@sriinnu/drishti/live.js";

// Launch the live TUI
await startLive();

import { runStatusline } from "@sriinnu/drishti/statusline.js";

// Run a single statusline tick
await runStatusline();

import { runDaemonCLI } from "@sriinnu/drishti/daemon/server.js";

// Control the daemon
await runDaemonCLI("start");
await runDaemonCLI("status");
await runDaemonCLI("stop");

Supported Editors

Claude Code, OpenCode, Codex CLI, Cursor, Windsurf, Zed, VS Code Copilot, and more. Run drishti editors to see the full list.

Author

Srinivas Pendela — @sriinnu

License

MIT