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/LastSearch-HQ/lastsearchWrote 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/lastsearch-hq/lastsearch/mcp-lastsearch)<a href="https://agentmods.dev/mcp/lastsearch-hq/lastsearch/mcp-lastsearch"><img src="https://agentmods.dev/badge/mcp/lastsearch-hq/lastsearch/mcp-lastsearch/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/lastsearch-hq/lastsearch/mcp-lastsearch"><img src="https://agentmods.dev/badge/mcp/lastsearch-hq/lastsearch/mcp-lastsearch.svg" alt="Reviewed on agentmods" width="80" height="20"></a>Grade A, and why
lastsearch 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 11d 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
{
"lastsearch": {
"command": "npx",
"args": [
"-y",
"lastsearch"
],
"env": {
"LASTSEARCH_API_KEY": ""
}
}
}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.
- 11d ago First seen · 12 lines scan A d5a4620bedd2
lastsearch is an MCP server published in the GitHub repository LastSearch-HQ/lastsearch (20 stars, last pushed 3d ago), licensed Apache-2.0. 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
statcite
Official economic statistics with citations: World Bank, IMF WEO, ECB. Verify claimed figures. Remote server at statcite.com.
ookcite
MCP server for citation validation, formatting, and collection management. Validates DOIs, formats references in 2900+ CSL styles, manages citation collections, and catches hallucinated citations. Runs locally from the @turtletech/ookcite-mcp npm package. Needs 1 environment variable to run.
ookcite-mcp
Catch fabricated citations: verify DOIs/ISBNs and format them in 2900+ CSL styles. Runs locally from the @turtletech/ookcite-mcp npm package. Needs 1 environment variable to run.
reddit-radar
MCP server to find Reddit threads where your product genuinely fits, reconstruct the conversation, and gate drafted replies against a claim boundary you define. Read-only: never posts, votes, or automates an account. Runs locally from the reddit-radar-mcp npm package. Needs 9 environment variables to run.
groundcheck
Verify claims, resolve FIGI identity, extract claims, and sign x402 delivery receipts over MCP. Runs locally from the groundcheck-mcp npm package. Needs 1 environment variable to run.
citestamp
Citation-health over a CC0 citation graph: what supports, refutes, or cites a paper. Read-only. Remote server at mcp.citestamp.com.