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 jlldavies/go4-llm-design-patterns --skill go4git clone --depth 1 https://github.com/jlldavies/go4-llm-design-patternsWrote 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/jlldavies/go4-llm-design-patterns/go4)<a href="https://agentmods.dev/skills/jlldavies/go4-llm-design-patterns/go4"><img src="https://agentmods.dev/badge/skills/jlldavies/go4-llm-design-patterns/go4/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/jlldavies/go4-llm-design-patterns/go4"><img src="https://agentmods.dev/badge/skills/jlldavies/go4-llm-design-patterns/go4.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 | $0.00141 | $0.02064 |
| Opus 5 | $0.00071 | $0.01032 |
| Sonnet 5 | $0.00028 | $0.00413 |
| Haiku 4.5 | $0.00014 | $0.00206 |
Grade A, and why
go4 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.
How it starts
The opening of the file, as written. The whole thing — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GO4 — LLM engineering design patterns
94 patterns across seven categories, grounded in transformer mechanics. Two jobs: pick the right pattern, and get its token economics right — cost is not a fixed tier, it's a function of how the system is used, and the choice compounds over every call. Don't read the whole catalog — the value is the decision guides and the conflict graph.
Resolve the catalog root
GO4 = $GO4_ROOT if set → else ../../ from this skill file → else ask. Verify: ls $GO4/patterns/CONFLICTS.md.
Before a big build — ask, then isolate
- Ask before you plan. Regime (Step 2) and pattern choice turn on facts usually in the user's head: scale / call frequency (once? thousands/day? millions?), cost of being wrong (reversible, or irreversible and expensive to redo?), latency / budget / model limits. If a substantial build hinges on numbers you'd otherwise guess, ask first — a plan for the wrong regime is worse than one question. Skip it for small or fully-specified tasks; don't interrogate.
- Then isolate. For a substantial or risky build, recommend the work happen in an isolated fork
(git worktree/branch, not base) so it can be merged or cleanly reverted. The fork-and-revert
mechanics are fiddly — use
superpowers:using-git-worktreesrather than hand-rolling git — and flag the merge-or-revert decision up front so work isn't stranded.
Step 1 — Classify (most designs touch two or three)
| Category | Governs | Reach for it when |
|---|---|---|
| Signal | Prompt shaping | Output format, personas, constraints, few-shot |
| Knowledge | Context engineering | RAG, retrieval, memory, compression, long context |
| Reasoning | Thinking structure | CoT, ReAct, tool loops, self-consistency, reflection |
| Orchestration | Multi-agent coordination | Pipelines, routing, parallelism, subagents, hierarchies |
| Reliability | Production safety | Bounds, logging, evals, human oversight, injection defence |
| Integration | Tool use | Function calling, MCP, CLI, agent-to-agent delegation |
| Humanizers | Cross-session continuity | Identity, persistent memory, self-improvement |
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 · 128 lines · 141 tokens per session scan A 6dd017756dcd
go4 is a skill published in the GitHub repository jlldavies/go4-llm-design-patterns (2 stars, last pushed 25d ago), licensed MIT. It adds 141 tokens to every session and 2,064 once invoked, about $0.0007 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
dspy
DSPy: declarative LM programs, auto-optimize prompts, RAG.
ai-engineering-toolkit
6 production-ready AI engineering workflows: prompt evaluation (8-dimension scoring), context budget planning, RAG pipeline design, agent security audit (65-point checklist), eval harness building, and product sense coaching.
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.
llm-ops
LLM Operations -- RAG, embeddings, vector databases, fine-tuning, prompt engineering avancado, custos de LLM, evals de qualidade e arquiteturas de IA para producao.
software-ai-integration
Applies production AI integration patterns for chat, structured output, guardrails, provider routing, and AI UX. Use when adding LLM-powered features to an application.
fine-tune-readiness
Decide whether fine-tuning is justified versus prompting or RAG, and gate the training dataset before any LoRA/SFT/DPO run. Use when someone says 'let's fine-tune'.