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 hajekim/agentic-design-patterns-skills --skill appendix-reasoning-enginesgit clone --depth 1 https://github.com/hajekim/agentic-design-patterns-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/hajekim/agentic-design-patterns-skills/appendix-reasoning-engines)<a href="https://agentmods.dev/skills/hajekim/agentic-design-patterns-skills/appendix-reasoning-engines"><img src="https://agentmods.dev/badge/skills/hajekim/agentic-design-patterns-skills/appendix-reasoning-engines/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/hajekim/agentic-design-patterns-skills/appendix-reasoning-engines"><img src="https://agentmods.dev/badge/skills/hajekim/agentic-design-patterns-skills/appendix-reasoning-engines.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 | $0.00397 | $0.03385 |
| Opus 5 | $0.00198 | $0.01692 |
| Sonnet 5 | $0.00079 | $0.00677 |
| Haiku 4.5 | $0.00040 | $0.00338 |
Grade A, and why
appendix-reasoning-engines 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 9d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- appendix-reasoning-engines — 100% identical, 3 lines differ
How it starts
The opening of the file, as written. The whole thing — 334 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Appendix F — Reasoning Engines: Under the Hood
Overview
Modern LLMs can be divided into two broad categories based on how they generate output:
- Standard Models: Generate tokens directly from the prompt context. Fast and cost-efficient.
- Reasoning Models: Perform extended internal deliberation — generating hidden "thinking" tokens before producing the final answer. Slower and more expensive, but significantly better at complex multi-step problems.
Understanding which type to use, and how they work internally, is critical for designing effective agentic systems.
Core Principle: Match the reasoning depth to the task complexity — reasoning models excel where accuracy matters more than speed; standard models win on throughput and cost.
When This Skill Applies
Activate this skill when:
- A task requires complex multi-step mathematical, logical, or strategic reasoning
- Accuracy is more important than latency (e.g., medical, legal, financial decisions)
- Standard models fail repeatedly on a problem that requires deliberate thinking
- You need to choose between
gemini-2.5-flash,gemini-2.5-flashwith high Thinking Budget, orgemini-2.5-pro - Building agents that need to plan deeply before acting
DEFINE → PLAN → ACTION Workflow
DEFINE
Assess the reasoning requirements:
- How many reasoning steps does the task require?
- Is accuracy or speed the primary constraint?
- What is the token budget and cost tolerance?
- Does the problem require backtracking or hypothesis testing?
PLAN
Select the right reasoning configuration:
- Map tasks to Standard vs. Reasoning model based on complexity
- Design the "thinking budget" — how many thinking tokens to allow
- Identify where to surface reasoning traces for debugging or auditing
- Plan cost controls (thinking tokens cost more than output tokens)
ACTION
Implement the reasoning-optimized agent:
- Use thinking-enabled models for complex planning and reasoning nodes
- Use standard models for simpler steps (retrieval, formatting, routing)
- Expose thinking traces for transparency when required
- Benchmark latency and accuracy trade-offs against your SLAs
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.
- 9d ago First seen · 334 lines · 397 tokens per session scan A 06ff00234e79
appendix-reasoning-engines is a skill published in the GitHub repository hajekim/agentic-design-patterns-skills (4 stars, last pushed 5mo ago), licensed MIT. It adds 397 tokens to every session and 3,385 once invoked, about $0.0020 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
prompt-optimizer
Analyze raw prompts, identify intent and gaps, match ECC components (skills/commands/agents/hooks), and output a ready-to-paste optimized prompt. Advisory role only — never executes the task itself. TRIGGER when: user says "optimize prompt", "improve my prompt", "how to write a prompt for", "help me prompt", "rewrite…
gemini-api
Google Gemini API patterns for Python and TypeScript. Covers content generation, streaming, tool use (function calling), vision, system instructions, context caching, batch requests, and agent workflows. Use when building applications with the Gemini API or Google Generative AI SDKs.
image-prompt
A Korean-language skill that turns a rough image idea into a detailed prompt for gpt-image-2, OpenAI’s image-generation model.
cost-aware-llm-pipeline
Cost optimization patterns for LLM API usage — model routing by task complexity, budget tracking, retry logic, and prompt caching.
data-scraper-agent
Build a fully automated AI-powered data collection agent for any public source — job boards, prices, news, GitHub, sports, anything. Scrapes on a schedule, enriches data with a free LLM (Gemini Flash), stores results in Notion/Sheets/Supabase, and learns from user feedback. Runs 100% free on GitHub Actions. Use when…
pytorch-patterns
PyTorch deep learning patterns and best practices for building robust, efficient, and reproducible training pipelines, model architectures, and data loading.