# AI & LLM, forgehouse

> Build with Claude the way we do. Prompt design, agent coordination, quality evaluation and the working habits that keep an AI model on task.

This is our discipline for working with Claude, written down: inputs that are designed instead of guessed, agents that are measured instead of taken on faith, and a memory layer that survives between sessions. It is the same method behind our own systems.

## Who this shelf is for
- Developer whose agents work in demos and break in production
- Team building on Claude with no way to measure output quality
- Builder who needs the AI to remember context after the session ends

## Where to start
Brain ships as its own product, so start here with Context Driven Development to fix what you feed the model, then add Agent Eval Suite Langsmith to measure what changes.

## Skills (17)
- [Agent Eval Suite Langsmith](https://forgehouse.ai/skills/agent-eval-suite-langsmith/), Production agent eval suite LangSmith dataset curation + Promptfoo assertion framework +…
- [Brain Context Engineering](https://forgehouse.ai/skills/brain-context-engineering/), Engineer what goes into an AI agent's context window: how much, in what order, and how compressed.
- [Brain Memory Hybrid Search](https://forgehouse.ai/skills/brain-memory-hybrid-search/), Bir agent'in memory/bilgi korpusu icin BM25 (lexical) + pgvector (semantic) hibrit arama, RRF skor birlesimi…
- [Claude Agent Template Library](https://forgehouse.ai/skills/claude-agent-template-library/), A categorized canon of 100+ Claude Code subagent templates with a strict frontmatter standard, enforcing Mission Brief, Agent Chain and bypass-permissions discipline.
- [Context Driven Development](https://forgehouse.ai/skills/context-driven-development/), When working with Conductor's context-driven development methodology, managing project context…
- [GPT & Perplexity Prompt Trace](https://forgehouse.ai/skills/gpt-perplexity-prompt-trace/), AI Search Forensic Intelligence
- [Hybrid Search Implementation](https://forgehouse.ai/skills/hybrid-search-implementation/), Combine vector and keyword search for improved retrieval.
- [Langchain Architecture](https://forgehouse.ai/skills/langchain-architecture/), Design LLM applications using LangChain 1.x and LangGraph for agents, memory, and tool…
- [LLM Evaluation](https://forgehouse.ai/skills/llm-evaluation/), Implement comprehensive evaluation strategies for LLM applications using automated metrics…
- [LLM Fine-Tuning Pipeline](https://forgehouse.ai/skills/fine-tuning-pipeline-llm/), spesifik LLM uretmek icin uctan uca fine-tuning playbook OpenAI hosted FT (GPT-4o-mini/4.1)…
- [ML Pipeline Workflow](https://forgehouse.ai/skills/ml-pipeline-workflow/), Build end-to-end MLOps pipelines from data preparation through model training, validation, and…
- [Model Selection Router](https://forgehouse.ai/skills/model-selection-router/), Lock every AI call to the most capable Opus model (no downgrade) and route cost savings through prompt caching, batch APIs and context engineering instead of cutting quality.
- [Multi Agent Orchestration Langgraph](https://forgehouse.ai/skills/multi-agent-orchestration-langgraph/), LangGraph ile production multi-agent orkestrasyon state machine (nodes + edges + state)…
- [Prompt Architect](https://forgehouse.ai/skills/prompt-architect/), AI image and video prompt builder that guides users from a rough vision to a copy-paste ready…
- [Prompt Caching Optimizer](https://forgehouse.ai/skills/prompt-caching-optimizer/), a brand prompt caching API ile %85-90 token maliyeti azaltma stratejisi.
- [Prompt Engineering Patterns](https://forgehouse.ai/skills/prompt-engineering-patterns/), Master advanced prompt engineering techniques to maximize LLM performance, reliability, and…
- [Voice AI Agent Vapi](https://forgehouse.ai/skills/voice-ai-agent-vapi/), Vapi.ai + Bland.ai + Retell AI sesli AI agent kurulumu

## Agents (1)
- [Skill Alchemist](https://forgehouse.ai/agents/skill-alchemist/), Skill upgrader to Ultra (Pro+) standard

## FAQ

### I am new to Claude Code, is this the right shelf to start?
It is the shelf about working with Claude itself. Claude Agent Template Library and Context Driven Development teach the structures; if you want a packaged starting point instead, the Brain ships the whole system pre-wired.

### What is context engineering and why does it matter?
An LLM is only as good as what it holds in view. These skills encode how we keep Claude on task across long work: what loads when, what stays out, how memory and rules persist, the difference between a demo and a system.

### Can I measure whether my agents are actually good?
Yes: Agent Eval Suite Langsmith builds evaluation runs for your agents: test cases, scoring, regression checks. You stop judging agents by vibe and start judging them by results.

## Related topics
- [Development](https://forgehouse.ai/catalog/development/), 58 pieces
- [Automation & Ops](https://forgehouse.ai/catalog/automation-ops/), 22 pieces
- [Data & Analytics](https://forgehouse.ai/catalog/data-analytics/), 27 pieces

https://forgehouse.ai/catalog/ai-llm/