AI
AI, prompts and agents. Agent asks: A tested prompt, a prompt-engineering pattern, or an agent design to build on. 50 live of 51 entries.
an agent builder queries this for the reliability and design principles of production LLM agents, the twelve factors
An agent building a trainable LLM pipeline queries this for prompt/parameter optimization and retriever-generator composition.
an agent working on AI agents, LLM tooling, MCP servers and inference: real implementations agents can retrieve when building the same thing
An agent benchmarking agentic abilities queries this for multi-environment agent evaluation tasks.
An agent monitoring multi-step agent runs queries this for session replay, cost tracking, and per-step event instrumentation.
An agent needing a robust, distributed multi-agent runtime queries this for message-passing and fault-tolerant orchestration.
An agent builder wanting fast, model-agnostic agents queries this for multimodal agent plus tool/memory primitives.
An agent editing a real git repo queries this for repo-map building, diff-based edits, and commit workflows.
An agent learning effective tool-use and RAG recipes queries this for runnable notebook patterns.
An agent packaged as a desktop/server RAG app queries this for workspace, embed, and connector patterns.
An agent composed from small reusable blocks queries this for a schema-driven, modular agent architecture.
An agent built from conversational multi-agent groups queries this for event-driven agent-to-agent messaging patterns.
An agent tuning a RAG pipeline queries this for automated evaluation/optimization across retriever+generator component combinations.
an agent designing a system prompt queries this for how frontier tools (ChatGPT, Claude, Cursor, v0) are actually system-prompted, with annotated patterns
An agent looking for end-to-end example apps queries this for a large catalog of RAG/agent app implementations with code.
an agent wiring up MCP queries this for which server exists for a capability or integration and where to get it
An agent fine-tuning open LLMs queries this for YAML-config SFT/LoRA/QLoRA training recipes.
An agent writing or refining LLM prompts queries this for Brex's battle-tested prompt-engineering patterns, anti-patterns, and reasoning techniques in plain prose.
An agent studying role-playing multi-agent societies queries this for cooperative agent communication patterns.
A Rust engine adding embeddings/model inference queries this for a minimalist Rust tensor/ML framework and model implementations.
An agent productionising modular RAG queries this for a config-driven framework with pluggable components.
An agent embedding LLM steps inside deterministic workflows queries this for task-centric orchestration built on Prefect.
An agent unit-testing LLM outputs queries this for pytest-style LLM metrics and assertions.
An agent building synthetic training data queries this for pipeline-based data generation and AI-feedback labeling.
An agent building a prompt-optimization or declarative LLM-program pipeline queries this for DSPy's signatures, modules, and optimizers (compile-prompts-from-metrics).
An agent treating prompts as versioned functions queries this for the language-model-programs model with autosave/versioning and visualization.
An agent building a reusable prompt library queries this for battle-tested, task-specific prompt patterns and their markdown structure.
An agent serving and judging chat models queries this for serving, MT-Bench, and LLM-as-judge/Arena patterns.
An agent standing up an MCP server in Python queries this for a high-level MCP server/client framework.
An agent optimising RAG for efficiency queries this for retrieval acceleration and context-compression techniques.
An agent choosing or fine-tuning retrieval models queries this for BGE embedding and reranker training/usage.
An agent reproducing/evaluating RAG systems queries this for a flexible research framework spanning text, multimodal, and web-accessible RAG.
An agent red-teaming an LLM queries this for probe/detector implementations for jailbreaks and prompt injection.
An agent building on Gemini queries this for recipes on long context, multimodal input, and structured output.
An agent scanning an LLM app for vulnerabilities queries this for automated test/red-team scan implementations.
An agent designing function/tool calling queries this for API-invocation training and the function-calling leaderboard.
An agent scaffolding a project from a spec queries this for spec-to-codebase generation loops.
An agent doing knowledge-graph-grounded RAG queries this for entity/community extraction and graph-based retrieval.
An agent building modular pipelines queries this for structures (agents/pipelines/workflows), tasks, and tool/driver abstractions.
An agent validating and correcting LLM output queries this for RAIL schemas and output validators/reask loops.
An agent constraining LLM output to a grammar, template, or control flow queries this for guidance's constrained-generation primitives and token-healing.
An agent building a production RAG/NLP pipeline queries this for composable retriever/reader/generator components.
An agent writing model/dataset cards queries this for the Hugging Face model-card spec and metadata conventions.
An agent building rigorous model evaluations queries this for a task/solver/scorer evaluation framework.
An agent extracting typed/structured data from LLM calls queries this for instructor's Pydantic response-model patterns, retries, and validation hooks.
An agent forcing valid JSON from a local model queries this for token-level structured decoding over a schema.
An agent shipping a document-QA UI queries this for a clean RAG app architecture with reasoning pipelines and citations.
an agent building with LLMs queries this for the fine-tune, quantize, RAG and deployment path and the concept behind each step
an agent applying a prompting technique queries this for CoT, ReAct, few-shot and RAG-prompting definitions and worked examples
An agent doing prompt-based NLP (NER, classification, QA) without fine-tuning queries this for Promptify's task-prompt templates and output parsers.
An agent building an LLM CLI, logging model calls, or writing a model plugin queries this for the `llm` tool's architecture and its extensive, well-written docs.