What is web-search-plus?

A web research tool set that searches the internet and extracts text from specific pages, returning sources rather than ready-made answers. It supports targeted research such as documentation, papers, security advisories, news, and release information.

Use it for source-backed web searches, reading long web pages, API and documentation lookups, arXiv papers, CVEs, open-source discovery, shopping, and current events.

It helps when search results need source details, a particular provider, a locale, or the full text of a page. The agent must read and combine the returned sources itself.

web-search-plus — Skill for Claude Code and Codex by robbyczgw-cla

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.

Any agent

npx skills add robbyczgw-cla/web-search-plus-mcp --skill web-search-plus

Clone the repo

git clone --depth 1 https://github.com/robbyczgw-cla/web-search-plus-mcp

Made for: Claude Code, Codex.

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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/robbyczgw-cla/web-search-plus-mcp/web-search-plus/github.svg)](https://agentmods.dev/skills/robbyczgw-cla/web-search-plus-mcp/web-search-plus)

Your own site

<a href="https://agentmods.dev/skills/robbyczgw-cla/web-search-plus-mcp/web-search-plus"><img src="https://agentmods.dev/badge/skills/robbyczgw-cla/web-search-plus-mcp/web-search-plus/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/robbyczgw-cla/web-search-plus-mcp/web-search-plus"><img src="https://agentmods.dev/badge/skills/robbyczgw-cla/web-search-plus-mcp/web-search-plus.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 132 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,450 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.

Model	Per session	Once invoked
Fable 5.1	$0.00132	$0.01450
Opus 5.5	$0.00053	$0.00580
Sonnet 5.5	$0.00026	$0.00290
Haiku 4.5	$0.00013	$0.00145

Measured 5d ago against content hash 90cd524b084d, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

web-search-plus 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 5d 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.

skills/web-search-plus/SKILL.md · 135 lines

How it starts

The opening of the file, as written. The whole thing — 135 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Web Search Plus

Web Search Plus gives you two MCP tools. Both return sources, never answers. You read the sources and synthesize; the server does not generate truth claims.

web_search — multi-provider search with automatic routing. Returns ranked results with URL, title, snippet, provider, and routing metadata.
web_extract — fetches one or more URLs and returns bounded markdown (or HTML), with optional query-ranked spans for long pages.

Both tools return compact JSON by default. Use detail: "full" when you need the complete v3 contract, observations or routing receipts. The server default can also be set with WEB_SEARCH_PLUS_MCP_DETAIL or config mcp.detail; an explicit tool argument wins.

When to prefer these over the built-in search

The user wants to see which sources say something, or wants a quality/routing report.
The query is provider-shaped: docs/API lookups, arXiv, CVE/security, GitHub/OSS discovery, local/shopping, news with a recency window.
The user wants a specific provider or a specific locale.
You need the full text of a page, not a snippet. Use web_extract.
The user says "wsp", "web search plus", or names a provider.

For quick, one-line factual lookups the built-in search is fine. Do not run both for the same question unless the user asks to compare.

`web_search`

Required: query.

Useful parameters:

provider — auto (default) or one of serper, brave, tavily, linkup, exa, firecrawl, parallel, you, searxng, keenable, serpbase, querit, donsetch, octen, tinyfish. Only providers with a configured key work; leave auto unless the user asks.
count — results, default 5, max 20.
freshness — day, week, month, year for recency.
search_type — news for a news vertical.
country / language — locale overrides.
include_domains / exclude_domains — allow or deny lists.
mode — research queries up to three providers concurrently, dedupes, and extracts the top sources. Slower and costs more; use for deep questions.
quality_report — true with detail: "full" to include routing class, provider scores, and authority signals. Useful when the user asks why a result ranked where it did.

Read the full file on GitHub · 135 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.

5d ago Changed 90cd524b084d
6d ago Changed · +5 lines df0f893d3bec
18d ago Changed · +7 lines 0bdfe420866b
26d ago Changed 10cd595169e2
27d ago Changed 534fbe3f6ea2
28d ago Changed 228dbcf5c7cf
29d ago Changed 8d781e23c2d8
1mo ago Changed bfcf8a569a68
1mo ago Changed 5a4704dfc733
1mo ago First seen · 123 lines · 132 tokens per session scan A 4e62c98004c2

Subscribe to this mod's changes

web-search-plus is a skill published in the GitHub repository robbyczgw-cla/web-search-plus-mcp (5 stars, last pushed 6d ago), licensed MIT. It adds 132 tokens to every session and 1,450 once invoked, about $0.0005 per session on Opus 5.5. 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 2026-09-05.

Python Research ai-agents donsetch hermes-agent mcp research-agent

web-search-plus next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
web-search-plusrobbyczgw-cla/web-search-plus-mcp	Skill	5	6d ago	MIT	A	132	Claude Code, Codex
repo-readiness-auditasimons81/hermes-field-kit	Skill	127	11d ago	Apache-2.0	A	48	Claude Code, Codex
dont-lie-to-measimons81/hermes-field-kit	Skill	127	11d ago	Apache-2.0	A	43	Claude Code, Codex
x-post-writerasimons81/hermes-field-kit	Skill	127	11d ago	Apache-2.0	A	46	Claude Code, Codex
x-analytics-importasimons81/hermes-field-kit	Skill	127	11d ago	Apache-2.0	A	32	Claude Code, Codex
research-intelligenceAlekseiUL/hermes-researcher-agent	Skill	65	8d ago	MIT	A	41	Claude Code, Codex
python-debugpyEKKOLearnAI/hermes-studio	Skill	12k	5d ago	none	A	26	Claude Code, Codex