Atlas / MCP servers / cavi-ai / mcp-eval

mcp-evalSAFE

mcp/cavi-ai/mcp-eval

🧐 Privacy-preserving MCP server evaluation: deterministic CI probe battery (18 probe kinds, readiness score, SARIF) plus production friction capture with findings that verify and close

Verdict
SAFE
Grade
B
Trust score
89 /100
Exposed tools
1 1r · 0w · 0d
Transport
stdio · streamable-http
License
MIT
Stars
2
01

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

[](LICENSE)

mcp-eval is an agnostic evaluation tool for MCP servers, with two complementary tracks:

  1. Deterministic benchmark battery — run a strict, manifest-driven probe

suite against any MCP server (stdio or Streamable HTTP) and get a pass/fail verdict with fixed failure reasons. No LLM in the loop, no network services, fully reproducible: same server, same manifest, same result. Built for CI gating while you develop a server.

  1. Friction capture — run your server behind a privacy-preserving shim in

production-like agent sessions, index what it actually costs agents, and promote repeated friction into queryable findings a development agent can act on. Findings close only after three consecutive green verification runs and reopen automatically on regression.

Capture journals and probe summaries omit raw payloads, error prose, and credentials. Store records under /store/ are eligible for sharing after reviewing annotation notes and the remaining files. Manifests, corpus target files, and retained corpus evidence are private artifacts outside that boundary; see Privacy and authorization.

Documentation: cavi-ai.xyz/docs/mcp-eval.

Install

brew install cavi-ai/tap/mcpeval
npm install --global @cavi-ai/mcp-eval
cargo install mcpeval --locked

Each installs mcpeval and the demo server mcpeval-demo. Homebrew (macOS and Linux, x64/arm64) and npm (also Windows x64) install the prebuilt release archive after checking its pinned SHA-256; the archives and their checksums are also on the releases page. Building from a checkout works too:

cargo build --release

The binaries are target/release/mcpeval and target/release/mcpeval-demo (.exe on Windows).

Quick

Read from source at commit fce0d5110a8eOBSERVED · 2026-10-11
02

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.

claude-code
claude mcp add mcp-eval -- npx -y @cavi-ai/[email protected]
claude-desktop
{
  "mcpServers": {
    "mcp-eval": {
      "command": "npx",
      "args": [
        "-y",
        "@cavi-ai/[email protected]"
      ]
    }
  }
}
03

Exposed tools (1)

1 read · 0 write · 0 destructive.

ToolRiskDescription
read_statereadRead the fixture dataset
04

Trust audit

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

LayerWhat it checksResult
L0Provenance & inventoryPASS
L1Static analysis of the codePASS
L2Instruction surface (what it tells the agent)PASS
L3Class-specific surfacePASS
L4Behavioural (sandbox)SKIPPED

What the source does

Filesystem
declared (1 observation(s))
Network
declared (6 observation(s))
Shell
declared (1 observation(s))
Dependencies
pinned
Secrets in source
none-found

Findings (12)

LOWInventory / provenance · inv.hidden_file · CWE-1104
.pr-gates
.pr-gates
Why it matters. hidden member outside the usual dotfiles
Fix. review its purpose
LOWFilesystem / path · fs.traversal · CWE-22, CWE-59
actions/corpus-evidence/index.mjs:2
import { checkEvidence } from "../../scripts/corpus/check-evidence.mjs";
LOWFilesystem / path · fs.traversal · CWE-22, CWE-59
scripts/corpus/contract.test.mjs:7
const corpusFixture = JSON.parse(await readFile(new URL("../../tests/fixtures/corpus-v3.json", import.meta.url), "utf8"));
LOWFilesystem / path · fs.traversal · CWE-22, CWE-59
scripts/corpus/contract.test.mjs:8
const metadataCases = JSON.parse(await readFile(new URL("../../tests/fixtures/corpus-v3-metadata-cases.json", import.meta.url), "utf8"));
LOWFilesystem / path · fs.traversal · CWE-22, CWE-59
scripts/corpus/deployment.test.mjs:50
for (const change of [{ root: "relative" }, { sha256: "unknown" }, { executable: "../outside" }, { executable: "/outside" }, { entrypoint: "a/../../outside" }, { executable: "runtime\\outside" }]) {
LOWFilesystem / path · fs.traversal · CWE-22, CWE-59
scripts/distribution/publish-auth.test.mjs:11
new URL("../../.github/workflows/publish.yml", import.meta.url),
LOWNetwork egress · net.raw_ip · CWE-200, CWE-319
README.md:168
--probe discovery-cost --url http://127.0.0.1:8080/mcp
LOWNetwork egress · net.raw_ip · CWE-200, CWE-319
README.md:393
--listen 127.0.0.1:8090 --upstream http://127.0.0.1:8080/mcp
LOWNetwork egress · net.raw_ip · CWE-200, CWE-319
docs/ci.md:90
url: http://127.0.0.1:8080/mcp
LOWNetwork egress · net.raw_ip · CWE-200, CWE-319
docs/install.md:257
--url http://127.0.0.1:8080/mcp
LOWNetwork egress · net.raw_ip · CWE-200, CWE-319
docs/mcp-eval/source/pages/guides/capture-and-probe.md:20
--upstream http://127.0.0.1:8080/mcp
LOWPrompt injection · prompt.transfer_instruction · CWE-94, CWE-1427
docs/mcp-eval/source/pages/guides/readiness-reporting.md:96
`--format markdown` renders a pull-request-ready report: the readiness score with its standard and a static shields.io badge URL encoding only the score, an area table, the surface, *Lost points* with
Why it matters. an instruction to move sensitive data to an outside destination
Fix. remove; a skill never needs the user's secrets off the machine

Gates applied: no_behavioural_pass.

Audited 2026-10-11 · audit v0.4.1 · source sha fce0d5110a8efull audit observations/trust-audit/mcp-server/cavi-ai__mcp-eval.json · Report an issue / request a re-scan
05

Audit history

Every audit this server has had. A grade with a past is a grade somebody is still checking.

DateSourceVerdictGradeScoreChange
2026-10-11fce0d5110a8eSAFEB89first audit
06

Questions

What is the mcp-eval MCP server?

🧐 Privacy-preserving MCP server evaluation: deterministic CI probe battery (18 probe kinds, readiness score, SARIF) plus production friction capture with findings that verify and close

What tools does mcp-eval expose?

1 in total: 1 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.

Is mcp-eval safe to connect to an agent?

The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.

What credentials does mcp-eval need?

No credential environment variables were found in its source, so it appears to need none.

How does mcp-eval run?

It speaks stdio and streamable-http, so it runs as a local process your client starts. It is published on npm as @cavi-ai/mcp-eval at 0.6.0.

How current is this page?

The grade is for one exact copy of the source (fce0d5110a8e), read on 2026-10-11. The repository is watched and re-audited when it changes.

Advertisement