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.
git clone --depth 1 https://github.com/sydasif/web-search-mcpWrote 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.
[](https://agentmods.dev/mcp/sydasif/web-search-mcp/web-research)<a href="https://agentmods.dev/mcp/sydasif/web-search-mcp/web-research"><img src="https://agentmods.dev/badge/mcp/sydasif/web-search-mcp/web-research/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.
<a href="https://agentmods.dev/mcp/sydasif/web-search-mcp/web-research"><img src="https://agentmods.dev/badge/mcp/sydasif/web-search-mcp/web-research.svg" alt="Reviewed on agentmods" width="80" height="20"></a>Grade A, and why
Web-Research 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 9d 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.
What it actually says
{
"Web-Research": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/sydasif/web-search-mcp.git",
"web-search-mcp"
],
"env": {
"EXA_API_KEY": "${EXA_API_KEY}",
"AUTH_TOKEN": "${AUTH_TOKEN}",
"CT0": "${CT0}"
}
}
}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.
- 9d ago First seen · 15 lines scan A 8c57c369ecd9
Web-Research is an MCP server published in the GitHub repository sydasif/web-search-mcp (24 stars, last pushed 5d ago), licensed MIT. Its token cost is not measured: an MCP server costs its tool schemas, not its config file. 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 2026-08-30.
Other mcp servers, from other repositories
research-automation-mcp-server
MCP server for autonomous research automation: web search, Wikipedia, arXiv, Crossref, fact-checking and report compilation. Runs locally from the research-automation-mcp-server Python package.
mcp-retrieval
MCP server "mcp-retrieval" as configured in Role1776/mcp-retrieval. Runs in Docker (ghcr.io/role1776/mcp-retrieval:0.1.7). Needs 23 environment variables to run.
obsidian-github-mcp
MCP server "obsidian-github-mcp" as configured in SarutobiSasuke8/obsidian-github-mcp. Runs locally from the @sarutobi-sasuke/obsidian-github-mcp npm package.
hn-tech-signal-mcp
MCP server for tech & AI intelligence – aggregates HackerNews, arXiv, Lobste.rs and GitHub. No API key required. Runs locally from the hn-tech-signal-mcp Python package.
reddit-agent-publisher
MCP server "reddit-agent-publisher" as configured in Bl0ck154/reddit-agent-publisher. Runs locally from the reddit-agent-publisher npm package.
codeofpaper
CLI for Code of Paper — discover GitHub implementations of research papers. Runs locally from the codeofpaper Python package. Needs 1 environment variable to run.