AI Agent Hub
Back to plugins
dsh-quota-meter-plus preview

dsh-quota-meter-plus

admin-security Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install Lstalu/dsh-quota-meter-plus

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install Lstalu/dsh-quota-meter-plus in your terminal to install the plugin; the source repository is available at https://github.com/Lstalu/dsh-quota-meter-plus

About this plugin

When you drive DeepSeek API calls through DSH, a multi-turn chat or an unbounded sub-agent chain can quietly accumulate significant token spend with zero real-time cost feedback in the interface. dsh-quota-meter-plus closes that gap with a complete per-session quota control loop: it bills on the real usage payloads returned by dsh llm/stream across cache-hit, cache-miss, and output tiers; renders a 2-pixel progress bar above the input that drains live; flashes a deduction badge on every charge; and hard-blocks new model calls the moment the budget runs out.

The pricing table is fully editable in the UI and persisted per model, with built-in peak/off-peak time-of-day switching (default peak: 09:00-12:00 and 14:00-18:00 Beijing time), weekend-wide off-peak pricing, and a fallback rate for unknown models. Sub-agent spend is traced up the agent chain and folded into the parent session quota, and auxiliary calls such as title generation and context compression are billed as well. All ledger data persists locally, survives DSH restarts, and is cleaned up when the session closes.

The plugin is built for independent developers, small teams, or course projects that rely on DSH plus the DeepSeek API and want to see, control, and cap every bit of token spend before it becomes a surprise on the bill.

Screenshots

Use Cases

  • Track and cap real-time token spend during multi-turn chats or sub-agent chains
  • Auto-switch between peak, off-peak, and weekend pricing based on actual usage
  • Block new model calls when the budget is exhausted to prevent surprise overcharges

Best For

  • Independent developers relying on DSH and the DeepSeek API
  • Small teams that need to control API costs
  • Course projects that want to manage token spend within a budget