AI Agent Hub
Back to plugins
🧩

dsh-elegent-balance-tracker

admin-security Updated 2026.08.27

Run the following command in DeepSeek Harness:

dsh plugin install X-avier-W/dsh-elegent-balance-tracker

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

Run dsh plugin install X-avier-W/dsh-elegent-balance-tracker in DeepSeek Harness to install; the source is at https://github.com/X-avier-W/dsh-elegent-balance-tracker, then restart dsh web and refresh the browser page.

About this plugin

Running long conversations in DeepSeek Harness often turns cost into a black box. dsh-elegent-balance-tracker surfaces two numbers right on the interface: the cumulative cost of the current session and the remaining account balance, both rendered with the official DeepSeek pricing rules so you never have to guess.

The cost line beneath the input box reads token usage from the session log, matches each message to the official peak or off-peak rate based on its Beijing-time timestamp, and refreshes as soon as the model finishes responding. The balance badge in the sidebar fetches the official /user/balance endpoint at startup, re-syncs every minute, and deducts real-time estimated spend across all local sessions between syncs. Since v0.2.0 the plugin also scrapes the official pricing page every 12 hours, so newly released models or adjusted rates take effect automatically without an upgrade.

Built for individual developers and small teams who run multi-round development or batch tasks in DeepSeek Harness daily. Everything stays local: the API key only talks to DeepSeek endpoints, session logs are parsed on the machine and never sent to any external service, and sub-agent sessions are tracked independently. If you care about data boundaries and want cost to be transparent, this plugin keeps your focus where it belongs: the work itself.

Use Cases

  • Track cumulative cost in real time during long conversations without manual token math
  • Run multiple sessions in parallel with live balance estimation to avoid overruns
  • New models automatically use official pricing without manual config changes

Best For

  • Individual developers running multi-round development in DeepSeek Harness
  • Small teams batching model tasks who need cost transparency
  • Users with strict data-boundary requirements who keep API keys local