ClaudeTokenSAP is a Claude Code plugin marketplace.
Its first plugin, Token Optimizer, improves how Claude Code handles vague build prompts, oversized file reads, and long sessions.
It is built to reduce context waste without making Claude feel dumber, more generic, or overly interruptive.
The core idea is simple:
- ask better questions before coding when the request is vague
- stop huge noisy reads before they flood context
- summarize useful file reads in a compact way
- preserve lightweight session memory around compaction
- log what happened so the workflow can be tuned over time
Why teams use it
Context efficiency comparison
| Workflow | What usually happens | Token impact |
|---|
| Vague build prompt | Claude starts coding too early, then backtracks | High waste |
| Token Optimizer enabled | Claude asks focused clarifying questions first | Lower waste |
| Large noisy log read | Raw bulk floods context | Very high waste |
| Token Optimizer enabled | Full read is blocked and narrowed | 50–80% lower in repetitive cases |
| Large JSON read | Full structure gets dumped into session | High waste |
| Token Optimizer enabled | Shape, keys, counts, and summaries are used first | 40–75% lower in good-fit cases |
| Long coding session | Context drifts after compaction | Hidden waste |
| Token Optimizer enabled | Lightweight continuity is preserved | Better session quality |
Approximate token-saving ranges
| Input type | Typical optimization | Approximate reduction |
|---|
| Repetitive logs | Collapse noise, surface real failures | 60–85% |
| Large JSON | Shape/key summary instead of full bulk | 40–75% |
| Repeated file reads | Prevent duplicate noisy context | 20–50% |
| Vague build starts | Clarify before scaffolding | Often large downstream savings |
| Search-heavy sessions | Compress broad match output | 60–80% |
In highly repetitive cases, reductions can exceed 70%, but savings vary by workflow, file type, prompt quality, and session behavior.
Why this exists
Claude Code is strongest when context contains signal, not bulk.
In real development workflows, context gets wasted by things like:
- prompts such as
create a calculator
- large logs where only a few lines matter
- big JSON files where only the shape or a few keys matter
- repeated file reads that dump raw bulk into the session
- long conversations that lose thread after compaction
That creates three practical problems:
- Higher token usage
- Lower reasoning quality
- More time wasted clarifying build direction after code has already started
Token Optimizer is designed to intervene at exactly those points.
What’s new in v2.3.7
Version 2.3.7 makes onboarding questions fully dynamic. The hook now calls Claude Haiku to generate 3–7 custom questions tailored to the exact prompt — skipping what’s already obvious. "Create a React dashboard with PostgreSQL" won’t ask about framework or database. Falls back to the 11 static intent sets (v2.3.6) when the API is unavailable.
What’s new in v2.3.6
