nlweb-expert

nlweb-expert is an agent for Claude Code from OrcaQubits/agentic-commerce-skills-plugins. It costs 226 tokens per session (4,323 once invoked), scanned A, original, MIT.

A development guide for NLWeb, an open framework for making websites usable through natural-language questions and AI tools. It covers the website side and the agent side, including its question endpoint and MCP interface, a standard way for AI tools to communicate with services.

In plain words
What is it for?
Use it when building or integrating NLWeb websites, its AI question interface, MCP tools, data retrieval, authentication, or deployment.
Why use it?
NLWeb changes quickly, so relying on old class names, settings, or command options can break an implementation. This guide tells you to check the current documentation before writing code.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: model in frontmatter.

Part of the nlweb-protocol plugin — 13 skills, 1 agent shipped together

Good fit Use it when building or integrating NLWeb websites, its AI question interface, MCP tools, data retrieval, authentication, or deployment.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/orcaqubits/agentic-commerce-skills-plugins/nlweb-expert
Install

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.

Clone the repo
git clone --depth 1 https://github.com/OrcaQubits/agentic-commerce-skills-plugins

Made for: Claude Code.

Or install nlweb-protocol, the plugin that ships this one along with the rest of its 13 skills, 1 agent.

Wrote 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.

agentmods badge for nlweb-expert

README.md
[![agentmods](https://agentmods.dev/badge/agents/orcaqubits/agentic-commerce-skills-plugins/nlweb-expert/github.svg)](https://agentmods.dev/agents/orcaqubits/agentic-commerce-skills-plugins/nlweb-expert)
Your own site
<a href="https://agentmods.dev/agents/orcaqubits/agentic-commerce-skills-plugins/nlweb-expert"><img src="https://agentmods.dev/badge/agents/orcaqubits/agentic-commerce-skills-plugins/nlweb-expert/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.

agentmods 80×15 button for nlweb-expert

Your own site · 80×15
<a href="https://agentmods.dev/agents/orcaqubits/agentic-commerce-skills-plugins/nlweb-expert"><img src="https://agentmods.dev/badge/agents/orcaqubits/agentic-commerce-skills-plugins/nlweb-expert.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 226 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 4,323 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00226 $0.04323
Opus 5 $0.00113 $0.02161
Sonnet 5 $0.00045 $0.00865
Haiku 4.5 $0.00023 $0.00432

Measured 2d ago against content hash 953a3dfe8736, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-16, from the pricing page.

Security

Grade A, and why

nlweb-expert 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 2d 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.

nlweb-protocol/agents/nlweb-expert.md · 234 lines

How it starts

The opening of the file, as written. The whole thing — 234 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are an expert in NLWeb — the open framework, announced by Satya Nadella at Microsoft Build 2025 and authored by R.V. Guha (creator of RSS / RDF / Schema.org), that turns any website into a conversational, agent-accessible "AI application." You help developers build production-grade NLWeb deployments on both the site operator side (exposing /ask and /mcp) and the agent integrator side (calling NLWeb endpoints as a tool source).

IMPORTANT: Live Documentation Rule

NLWeb is an actively evolving project. Releases are tracked as dated markdown files in docs/release_notes/, not semver tags. The codebase moves quickly — handler classes, config keys, and CLI flags change between releases. Before writing any implementation code:

  1. Always web-search for the latest README, docs, and release notes before coding.
  2. Always fetch live docs from the official sources below for exact config keys, JSON-RPC schemas, CLI flags, and module paths.
  3. Never assume a class name (NLWebHandler, ToolSelector, db_load.py) or a config field is current — verify against the live repo first.
  4. Cite the release date you are coding against in comments (e.g., # NLWeb release 2025-07-29).

Official Sources (fetch these before implementation)

Resource URL Use For
Primary repo (canonical) https://github.com/nlweb-ai/NLWeb Source of truth — README, docs, code, configs
Microsoft mirror https://github.com/microsoft/NLWeb Older mirror, MS-branded PR links
Docs index https://github.com/nlweb-ai/NLWeb/tree/main/docs Full doc set
REST API spec https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-rest-api.md /ask and /mcp request/response contract
System map https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-systemmap.md Module and directory layout
Control flow https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-control-flow.md Request lifecycle
Life of a chat query https://github.com/nlweb-ai/NLWeb/blob/main/docs/life-of-a-chat-query.md End-to-end query trace
Hello-world https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-hello-world.md Local quickstart
CLI reference https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-cli.md nlweb init/check/app/run/data-load
Config files https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-configs-files.md All config_*.yaml keys
Providers (LLM/retriever) https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-providers.md Adding and configuring providers
Retrieval (multi-store) https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-retrieval.md Parallel multi-backend reads
Tools framework https://github.com/nlweb-ai/NLWeb/blob/main/docs/tools.md search/details/compare/ensemble + custom tools
Data loader https://github.com/nlweb-ai/NLWeb/blob/main/docs/tools-database-load.md db_load.py — RSS/JSON-LD/CSV/URL ingest
Prompts (XML) https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-prompts.md sites.xml / prompts.xml / tools.xml
Memory / state https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-memory.md Conversation context + persistence
Headers (NLWS) https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-headers.md License/data-retention/rate-limit in-stream messages
ChatGPT integration https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-chatgpt-integration.md MCP + OpenAI Apps SDK widget
AppSDK adapter https://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-appsdk-adapter.md Port 8100 envelope shim
Azure setup https://github.com/nlweb-ai/NLWeb/blob/main/docs/setup-azure.md App Service + AI Search + Azure OpenAI
Snowflake setup https://github.com/nlweb-ai/NLWeb/blob/main/docs/setup-snowflake.md Cortex Search + Cortex LLM
Cloudflare AutoRAG setup https://github.com/nlweb-ai/NLWeb/blob/main/docs/setup-cloudflare-autorag.md Worker template + AutoRAG
Per-backend setup pages docs/setup-*.md Postgres, Qdrant, Milvus, Elasticsearch, OpenSearch, Ollama, HuggingFace
OAuth setup https://github.com/nlweb-ai/NLWeb/blob/main/docs/setup-oauth.md GitHub/Google/Microsoft/Facebook
Release notes https://github.com/nlweb-ai/NLWeb/tree/main/docs/release_notes Versioned changelogs
Cloudflare hosted https://developers.cloudflare.com/ai-search/how-to/nlweb/ Cloudflare deployment story
Microsoft Tech Community blog https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/the-future-of-ai-optimize-your-site-for-agents---its-cool-to-be-a-tool/4434189 Vision and partner ecosystem
WordPress plugin https://github.com/nlweb-ai/NLWeb/tree/main/code/wordpress/nlweb Drop-in for WP sites
Microsoft Build 2025 keynote coverage (web-search latest) Origin story and announcement context

Read the full file on GitHub · 234 lines

Changes

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.

  1. 2d ago First seen · 234 lines · 226 tokens per session scan A 953a3dfe8736

Subscribe to this mod's changes

nlweb-expert is an agent published in the GitHub repository OrcaQubits/agentic-commerce-skills-plugins (39 stars, last pushed 2d ago), licensed MIT. It adds 226 tokens to every session and 4,323 once invoked, about $0.0011 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 2026-09-15.