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 lokikill123/codex-token-skills --skill token-savergit clone --depth 1 https://github.com/lokikill123/codex-token-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/lokikill123/codex-token-skills/token-saver)<a href="https://agentmods.dev/skills/lokikill123/codex-token-skills/token-saver"><img src="https://agentmods.dev/badge/skills/lokikill123/codex-token-skills/token-saver/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/lokikill123/codex-token-skills/token-saver"><img src="https://agentmods.dev/badge/skills/lokikill123/codex-token-skills/token-saver.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 | $0.00036 | $0.00279 |
| Opus 5 | $0.00018 | $0.00139 |
| Sonnet 5 | $0.00007 | $0.00056 |
| Haiku 4.5 | $0.00004 | $0.00028 |
Grade A, and why
token-saver 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 10d 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.
What it actually says
Token Saver + Cache Optimizer
Response Rules
- One sentence per response unless detail explicitly requested
- Never re-read files already read this turn
- Skip preambles, summaries, progress updates
- No update_plan unless explicitly requested
- No testing/building/formatting unless explicitly requested
- Patch files directly, announce: filename + 5-word description
- No Mermaid, no tables, no code blocks >10 lines
Cache Optimization (saves 50%+ token cost)
- Prompt caching is prefix-based: one byte change = entire cache miss
- Keep this file FROZEN - batch all changes, edit once
- Stable content first (rules, skills), dynamic content last (dates, file lists)
- Single skill file beats multiple (fewer reads = fewer break points)
- After editing: accept ~24h cache rebuild period
- Cache hit ≈ response <500ms, miss ≈ >1s first token
User Tips
- Batch tasks into single prompts
- Write output to disk instead of inline
- AGENTS.md for project context, not re-explaining
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.
- 10d ago First seen · 28 lines · 36 tokens per session scan A 3c53497a50f0
token-saver is a skill published in the GitHub repository lokikill123/codex-token-skills (111 stars, last pushed 2mo ago), licensed MIT. It adds 36 tokens to every session and 279 once invoked, about $0.0002 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 2026-08-30.
Other skills, from other repositories
cost-aware-llm-pipeline
A planning guide for choosing language models and managing the amount of conversation context used by an AI coding workflow. It groups tasks by complexity and gives rules for avoiding context overflow during long sessions.
prompt-engineering
Universal prompt engineering techniques for any LLM. Use when crafting, optimizing, or reviewing prompts for AI models. Triggers on requests like "improve this prompt", "write a system prompt", "optimize my instructions", "help me prompt engineer", "audit this prompt", "review my prompt", or when building agentic…
ask-short-answer-first
Master the Bottom Line Up Front (BLUF) prompting pattern to get the direct 2-line answer immediately, eliminating preamble fluff and saving reading time.
be-specific-instead-of-vague
How to replace vague prompts with the 5-Variable Precision Formula to eliminate generic fluff, hit exact requirements, and get one-shot perfection.
give-ai-direct-feedback
How to use the 3-Part Direct Feedback Formula to steer AI outputs without vague trial-and-error or starting chats from scratch.
give-ai-situation-context
How to anchor prompts with the 4-Pillar Situation Framework (Role, Scale, Budget, Constraints) to stop AI from over-engineering or giving irrelevant advice.