Awesome Reviewers expert instructions

Expert instructions for AI and infrastructure engineering

A library of instruction sets distilled from review discussions in production repositories — agent frameworks, inference servers, orchestration, databases, cloud infrastructure and the tooling around them. Each one is plain markdown: copy it here, or fetch it as raw text and load it into an agent, harness or context broker.

Use in an AI agent

Paste an instruction directly into your agent's system prompt. Agents (Claude, GPT, etc.) can apply domain-specific expertise on the fly, catching subtle violations that generic rules miss.

Use in a skill or plugin

Build a skill that loads instructions programmatically — e.g., /claude-code-review my-file.js fetches the relevant instruction for that language and applies it. Skills compose instructions with the target code and invoke the agent.

Use in a code review harness

Store instructions in a database or fetch from /raw/ endpoints. Gate them by language, repository, or domain. A review harness finds the matching instructions and passes them as context to the review model.

    Domains

    organized by engineering discipline

    Agent loops, tool calling, prompt and context assembly, agent SDKs and coding assistants.

    Inference servers, model gateways and routers, KV caching, retrieval and ingestion pipelines.

    Training and inference frameworks, tensor and kernel code, ML platforms and pipelines.

    Kubernetes, schedulers, service meshes, container tooling, workflow and job engines.

    Infrastructure as code, cloud SDKs, edge runtimes, proxies, gateways and tunnels.

    Query engines, storage, replication, streaming, ORMs and analytics backends.

    Metrics, tracing, logging, error tracking and instrumentation of production systems.

    Static analysis, cloud posture scanning, authentication, authorization and secrets handling.

    Interpreters, compilers, async runtimes, parsers and low-level systems code.

    Editors, terminals, build systems, package managers, linters, formatters and test runners.

    Web and mobile frameworks, component libraries, API layers and product codebases.

    Reference material, curated lists and learning resources — mostly writing and structure guidance.

    Recently updated

    browse all instructions

    • QwenLM/qwen-code Security TypeScript ai-agents
      raw open
    • QwenLM/qwen-code Testing TSX ai-agents
      raw open
    • QwenLM/qwen-code Testing JavaScript ai-agents
      raw open
    • QwenLM/qwen-code Documentation TypeScript ai-agents
      raw open
    • QwenLM/qwen-code Security Markdown ai-agents
      raw open
    • QwenLM/qwen-code Configurations TypeScript ai-agents
      raw open
    • QwenLM/qwen-code Testing Yaml ai-agents
      raw open
    • QwenLM/qwen-code Code Style TSX ai-agents
      raw open
    • QwenLM/qwen-code Testing TypeScript ai-agents
      raw open
    • looplj/axonhub Security Go llm-infra
      raw open
    • QwenLM/qwen-code API TSX ai-agents
      raw open
    • QwenLM/qwen-code API Yaml ai-agents
      raw open
    • QwenLM/qwen-code Logging Yaml ai-agents
      raw open
    • QwenLM/qwen-code CI/CD JavaScript ai-agents
      raw open
    • QwenLM/qwen-code Error Handling TypeScript ai-agents
      raw open
    • QwenLM/qwen-code Error Handling Yaml ai-agents
      raw open
    • nousresearch/hermes-agent Documentation Markdown ai-agents
      raw open
    • nousresearch/hermes-agent Security Python ai-agents
      raw open
    • nousresearch/hermes-agent Concurrency Python ai-agents
      raw open
    • QwenLM/qwen-code Security Yaml ai-agents
      raw open
    • QwenLM/qwen-code Code Style TypeScript ai-agents
      raw open
    • QwenLM/qwen-code CI/CD Yaml ai-agents
      raw open
    • maximhq/bifrost Concurrency Go llm-infra
      raw open
    • aaif-goose/goose Error Handling Rust ai-agents
      raw open
    • aaif-goose/goose API Rust ai-agents
      raw open

    Raw access

    endpoint reference

    # one instruction
    curl https://awesomereviewers.com/raw/qwen-code-secure-boundary-enforcement.md
    
    # every instruction in a domain, concatenated
    curl https://awesomereviewers.com/raw/bundles/ai-agents.md
    
    # machine index: slugs, domains, dates, raw URLs
    curl https://awesomereviewers.com/raw/index.json

    Stable paths, no key, no rate limit. Regenerated on every push to the corpus.