Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add tentenco/skills --skill token-saving-prompt-enhancergit clone --depth 1 https://github.com/tentenco/skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/tentenco/skills/token-saving-prompt-enhancer)<a href="https://agentmods.dev/skills/tentenco/skills/token-saving-prompt-enhancer"><img src="https://agentmods.dev/badge/skills/tentenco/skills/token-saving-prompt-enhancer/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/tentenco/skills/token-saving-prompt-enhancer"><img src="https://agentmods.dev/badge/skills/tentenco/skills/token-saving-prompt-enhancer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00241 | $0.01887 |
| Opus 5 | $0.00120 | $0.00944 |
| Sonnet 5 | $0.00048 | $0.00377 |
| Haiku 4.5 | $0.00024 | $0.00189 |
Grade A, and why
token-saving-prompt-enhancer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 150 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Token-Saving Prompt Enhancer
Design prompts for a Claude Code setup where the most expensive model only directs, and cheaper models do the work. A well-designed Enhanced Prompt can cut token spend dramatically; a naive one silently burns the orchestrator's context. Your job is to make the efficient version the default.
The target workflow you are designing for
The user's Claude Code is configured as:
| Role | Model | Duty |
|---|---|---|
| Orchestrator | Fable 5 (max reasoning) | Plan, decompose, synthesize. Context must stay lean — it is the most expensive model. |
| deep-reasoner subagent | Opus | Architecture, complex debugging, algorithm design. Returns concise conclusions only. |
| fast-worker subagent | Sonnet | Boilerplate, tests, formatting, simple multi-file edits. |
Codex (/codex:rescue --background) |
OpenAI Codex CLI | Peer senior engineer with a different perspective — a peer, not a reviewer. Costs zero Claude tokens. |
High-stakes (hard-to-reverse) decisions get the double-blind pattern: task deep-reasoner + Codex on the same problem in parallel, neither sees the other's answer, the orchestrator synthesizes the best of both.
If the user seems not to have this workflow set up yet, point them to
references/workflow-setup.md (or offer to walk them through it) before
designing prompts that assume it.
Input types you will receive
- A. Rough idea — "我想做一個…". Mostly underspecified; clarify first.
- B. Requirements — a feature list, bug description, or refactor goal.
- C. Prompt pain points — an existing prompt that wastes tokens, drifts off-goal, or produces bloated output. Diagnose before rewriting.
Five token-saving principles — encode ALL of them in every Enhanced Prompt
These are the reason this skill exists. A prompt missing any of them leaks tokens in a way the user won't notice until the bill arrives:
- The orchestrator never reads large files or logs itself. Subagents read and return summaries with file:line references. Reading a 2,000-line file in the orchestrator poisons its context for the rest of the session.
- Every delegated subtask specifies a return format with a size cap (e.g. "return: root cause + fix location, ≤10 lines"). Without a cap, Opus happily returns its whole chain of reasoning.
- Plan-first gate: require "Show me your plan first, then execute." A wrong direction caught at plan time costs hundreds of tokens; caught mid-execution it costs tens of thousands.
- Right-sized delegation: reasoning → deep-reasoner; mechanical → fast-worker; fresh perspective → Codex. Never send mechanical work to Opus (wasteful) or design decisions to Sonnet (unreliable).
- Verifiable success criteria so the orchestrator can loop and verify on its own instead of coming back to ask.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 150 lines · 241 tokens per session scan A 34b562f0a02b
token-saving-prompt-enhancer is a skill published in the GitHub repository tentenco/skills (4 stars, last pushed yesterday), licensed MIT. It adds 241 tokens to every session and 1,887 once invoked, about $0.0012 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
llm-app-patterns
Production-ready patterns for building LLM applications. Covers RAG pipelines, agent architectures, prompt IDEs, and LLMOps monitoring. Use when designing AI applications, implementing RAG, building agents, or setting up LLM observability.
prompt-optimization
Improve a prompt on the evaluations workbench through a measured loop. Score the baseline first, then duplicate the target column, form a hypothesis from failing rows, edit the copy's prompt draft, run, compare pass rate and cost, and repeat until the numbers hold. Use when the user asks to optimize or improve a…
enhance-prompt
Transforms vague UI ideas into polished, Stitch-optimized prompts. Enhances specificity, adds UI/UX keywords, injects design system context, and structures output for better generation results.
prompt-engineer
Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot…
seedance-vocab-en
This skill should be used when an English Seedance 2.0 prompt needs clearer production wording, less generic prose, or precise vocabulary for camera, lighting, motion, VFX, audio, and constraints. Route blocked prompts through seedance-filter for context and boundary review.
ideogram4
Prompting patterns for Ideogram 4 text-to-image — best-in-class in-image text rendering and exact color/layout control via structured JSON captions. Use when generating images that need legible on-image text (title cards, thumbnails, logos, signage, CTAs), precise brand colors, or controlled spatial layout. Triggers…