Agent-ready websites
Browse all WebMCP →Sites that expose structured tools agents can call — not just scrape.
Top servers this week
Top 100 · all 14,910 →Ranked by listing page views, last 7 days.
Categories
Full directory →How we verify
Methodology →
A handshake is a timed HTTPS request: initialize, then tools/list.
The verification grade adds that result to indexed tools, HTTPS or local transport, and a probe timestamp inside 7 days.
It is not a safety score. Tools are not called.
Discovery MCP and API
Use Influzer MCP Discovery in Claude, ChatGPT, or Cursor — search 14,910 servers listed by capability without leaving your chat.
https://www.influzer.ai/mcp/discovery
Connect Discovery · Install JSON · Probe badge · Agent skill
MCP Recipes
All recipes →Latest from Insights ORIGINAL
All articles →Around the web
community posts & press — updated dailySponsored · paid placement · handshake unchanged
2 slots open · $249/moNo VC · No tracking · No pay-to-rank
Influzer is run by one person and funded by its sponsors.
Every handshake, crawl, and index rebuild costs money. Two sponsor slots a month keep the registry free — and never touch the ranking.
- 65,721 visitors, last 30 days
- 175,414 listing page views, last 30 days
- 2 sponsor slots open
Top 100 handshake: 10 live · 40 auth · 83 HTTPS hosts · what changed
Put Influzer in the workflow
Search, handshake, paste, teach the agent — or watch the Top 100 clock. None of them is a safe-to-install badge.
/mcp/discovery
Snippets
Install JSON
Cursor, Claude, ChatGPT, Claude Code, ACP. Embed the widget on any docs page.
/api/v1/install/:slug
API
Handshake probe
Live tools/list for an HTTPS URL. CI on mcp.json, or a README SVG badge.
/api/v1/probe/badge
Skill
Search → paste
SKILL.md so the agent searches, handshakes, then pastes — it does not auto-connect.
/skills/influzer-mcp/SKILL.md
Live
Scoreboard
Handshake status for the Top 100. Sponsored pins are ads and do not change the row.
/mcp/scoreboard
Open source
ModelSpend
See LLM cost per project & agent, route to the cheapest suitable model, shadow savings. Star it on GitHub.
github.com/rmencke2/modelspend