Custom status-bar footer for the pi coding agent — model, thinking level, path, git branch, token usage, cost, context window usage, and cache hit rate, and generation throughput.

DSH usage & cost monitor: floating peak/off-peak aware instrument panel with official DeepSeek pricing benchmark, per-model cache hit rate, burn-rate forecast and cost calculator

Read-only performance diagnostics for DeepSeek Harness: session load (open/restore) timing, spill-hit counts, compaction count and trigger, context-injection volume (AGENTS.md/skills/tool-schema token share), and LLM cache hit rate — surfaced via the /fas

Per-session billing plugin for DeepSeek Harness: token buckets, cache hit rate, and peak-aware cost in the session header + a price-table settings page

Token usage & cost metering for dsh: per-session and aggregate token statistics (cache hit rate, input/output split) with configurable pricing, plus markdown reports.

DeepSeek balance & cache hit rate statusline for Claude Code

Floating draggable token usage statistics panel for DeepSeek Harness: totals, cache & hit rate, per-provider and per-model usage, current-month heatmap with per-day view, per-session detail and per-request records.

DSH plugin bundle: cumulative token usage, cache-hit rate, and per-model/per-period estimated cost, with a live session-header badge

Real-time status bar plugin for OpenCode: cache hit rate, token usage, cost, latency and model info in the sidebar footer.

DeepSeek LLM prompt-cache hit-rate optimizer for opencode, Claude Code, and Codex CLI.

DSH 上下文透镜 + 成本治理插件：token 用量可视化（每次请求的上下文构成、KV 缓存命中率）+ 每轮/每会话预算控制（超限提醒/拦截）。DSH web plugin: context lens (token usage visualization, KV cache hit rate) + cost governance (per-turn / per-session budget with warn/deny).

A generational cache with strict entry-count limits and payload validation.

Custom footer extension for pi — shows absolute token counts, cache hit rate, and thinking level

Codex-style context visibility for DeepSeek Harness: live context-window occupancy, composition breakdown, KV-cache hit rate, and a GitHub-style 53-week usage heatmap

Pi extension: Claude Code style HUD statusline — model, context usage, cost, cache hit rate, subagents, opencode-go quota

⚡ High-performance synthetic prompt generation with genetic evolution, streaming, and multi-model routing. 90%+ cache hit rate, <15ms P99 latency, no Redis required.

CCometixLine fork — high-performance Claude Code statusline, with prompt cache hit-rate in the context window segment and a second status line showing session/weekly usage.

Interactive TUI panel for OMP token usage analytics — multi-dimension grouping, cache hit rate, drill-down navigation, mouse support.

Accurate cache-hit rate, token usage, and cost estimates for the DeepSeek Harness Web UI: one-decimal hit rate, bucket breakdown, peak/off-peak pricing with auto price-switch, and a session usage panel.

Drop-in wrappers for the Anthropic, OpenAI (chat.completions + Responses), and Gemini SDKs (non-streaming AND streaming) that measure prompt cache hit rate, surface dollar savings, auto-manage Gemini CachedContent lifecycles with TTL auto-refresh, canonic