Snapshot of the MCP server ecosystem as of 2026-09-12 · by PulseFeed · see the live report →
354 of 3860 MCP servers ship an install/postinstall script — arbitrary code runs on your machine at npm i, before you've even used the tool. Some are legitimate native builds, but each is an unreviewed code-execution vector — and 422 of all audited servers publish no repository to inspect at all. A bad MCP server can read your files, env vars and keys the moment it's connected.
npm servers get deep signals; remote servers get liveness + HTTPS only.
| Server | Verdict | Flags |
|---|---|---|
| ai.agenticshelf/graffeo | avoid | unreachable |
| ai.agenticshelf/mcp | avoid | unreachable |
| ai.agenticshelf/puroair | avoid | unreachable |
| ai.alpic.test/test-mcp-server | avoid | unreachable |
| ai.autorfp/mcp | avoid | unreachable |
| ai.baselight/baselight | avoid | unreachable |
| ai.buyersense/buyersense | avoid | unreachable |
| ai.clarid/compliance | avoid | unreachable |
| ai.com.mcp/hapi-mcp | avoid | unreachable |
| ai.com.mcp/skills-search | avoid | unreachable |
| ai.dynamicfeed/dynamic-feed | avoid | unreachable |
| agentdb | avoid | install_script |
| Date | Audited | Install-script | Avoid | Abandoned |
|---|---|---|---|---|
| 2026-09-12 | 3860 | 354 | 365 | 131 |
| 2026-09-11 | 3836 | 353 | 364 | 131 |
| 2026-09-10 | 3793 | 351 | 362 | 129 |
| 2026-09-09 | 3701 | 347 | 358 | 129 |
| 2026-09-08 | 3640 | 345 | 356 | 129 |
| 2026-09-07 | 3545 | 344 | 355 | 128 |
| 2026-09-06 | 3538 | 343 | 354 | 128 |
| 2026-09-05 | 3534 | 344 | 355 | 127 |
| 2026-09-04 | 3517 | 344 | 355 | 127 |
| 2026-09-03 | 3490 | 342 | 353 | 124 |
| 2026-09-02 | 3456 | 340 | 351 | 124 |
| 2026-09-01 | 3350 | 333 | 344 | 122 |
| 2026-08-31 | 3299 | 332 | 343 | 122 |
| 2026-08-30 | 3272 | 330 | 341 | 121 |
Methodology: PulseFeed discovers MCP servers from the official MCP registry and npm, then audits each independently — install/postinstall scripts (code execution on install), abandonment, provenance, license, repository, download volume, and liveness for remote servers. Verdict = safe / caution / avoid. This report is generated automatically from that data and updates daily. Install scripts are not inherently malicious (native builds use them) but every one is an unreviewed code-execution vector worth checking. Same independent-audit approach as our x402 trust oracle.
Check any server free: GET /mcp/verify?package=<npm-name> · Machine-readable: /mcp-report.json · Live observatory: /mcp · /llms.txt