com.averray/mcp
version 1.1.0
- scanned 2026-10-16
- source github.com/averray-agent/agent
- commit 039764842f3a
- transport streamable-http
- files read 1397
- methodology 0.2.0
- ruleset 0.1.1
No tag matching version 1.1.0 was found in this repository, so the scan read the default branch instead. The findings below describe that code, not necessarily the released version.
68measured on 99% of this axis
66 findings in test code are listed but not deducted; 46 vendored or minified paths were excluded before scanning; machine-generated bundles are not the server's own code; 2 files could not be parsed and were not scanned
71measured on 85% of this axis
0
68measured on 55% of this axis
25measured on 80% of this axis
The evidence
Every deduction below names the artefact that produced it. Nothing on this page was executed: the source was read, never run, and no running server was contacted.
Code Safety
Showing the 50 most severe of 110 findings on this axis. The score was computed over all of them — findings beyond the fourth of a severity still deduct at a tenth of their weight — so the remainder is omitted from this page, not from the arithmetic.
- app/handoffs/averray-mobile-operator-control/project/support.js:71mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
const rest = {}; for (const k of Object.keys(obj)) { if (k[0] !== "$") rest[k] = obj[k]; } return { props: Object.keys(rest).length ? rest : null, preview }; - app/handoffs/averray-mobile-operator-control/project/support.js:189mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
for (const k in entry.propsMeta || {}) { const v = entry.propsMeta?.[k]?.default; if (v !== void 0) d[k] = v; } return d; - app/handoffs/averray-mobile-operator-control/project/support.js:1227mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
for (const k of Object.keys(window)) { if (!before.has(k) && typeof window[k] === "function") { globals[k] = window[k]; } } - mcp-server/src/core/agent-profile.js:305mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
} for (const key of Object.keys(totals)) { totals[key] += Number.isInteger(signals[key]) ? signals[key] : 0; } } - mcp-server/src/core/job-catalog-normalization.js:332mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
throw new ValidationError(`schedule.${key} must be ISO-8601 if provided`); } normalised[key] = new Date(parsed).toISOString(); } } - mcp-server/src/core/job-catalog-normalization.js:505mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
for (const key of ["pausedAt", "archivedAt", "reopenedAt"]) { if (typeof raw[key] === "string" && raw[key].trim()) { lifecycle[key] = normalizeIsoTimestamp(raw[key], `lifecycle.${key}`); } } - mcp-server/src/core/job-catalog-normalization.js:510mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
for (const key of ["reason", "staleReason"]) { if (typeof raw[key] === "string" && raw[key].trim()) { lifecycle[key] = raw[key].trim().slice(0, 500); } } - mcp-server/src/core/work-receipt.js:113mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
throw new ValidationError(`Verdict-core commitment requires root section ${section}.`); } verdictCore[section] = value; } return hashCanonicalContent(verdictCore); - mcp-server/src/services/arrival-dropoff.js:220mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
for (const actor of DROP_OFF_ACTORS) { for (const stage of ARRIVAL_STAGES) { totals[door][actor][stage] = topCodes(totals[door][actor][stage]); } } - mcp-server/src/services/arrival-observatory.js:886mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
} for (const source of ATTRIBUTION_SOURCES) { current.attributionSources[source] += Number(entry.attributionSources?.[source] ?? 0); } current.doors = [...new Set([ - mcp-server/src/services/arrival-observatory.js:1204mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
for (const source of ATTRIBUTION_SOURCES) { if (Number.isFinite(Number(stored?.[source]))) { target[source] = Number(stored[source]); } } - mcp-server/src/services/arrival-operator-view.js:202mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
current.doors.add(door); const reached = hasReachedSessionStage(session, "submitted") ? "submitted" : "claimed"; if (stageRank(reached) > stageRank(current.doorFurthest[door])) current.doorFurthest[door] = reached; } } - mcp-server/src/services/bootstrap.js:1606mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
]) { if (xcm[key] !== undefined) { normalized[key] = xcm[key]; } } - mcp-server/src/services/deposit-pool-observability.js:44mcp-js-prototype-pollution-merge-loop (medium/medium)
Keys from a caller-supplied object are copied onto a target without excluding __proto__ / constructor / prototype. A tool argument carrying __proto__ then mutates Object.prototype for the whole process, which can turn an unrelated later property read into attacker-controlled data.
}; for (const key of ["assetsRaw", "sharesRaw", "requestId", "tier", "unlockAt"]) { if (event[key] !== undefined && event[key] !== null) common[key] = String(event[key]); } return common; - app/handoffs/averray-mobile-operator-control/project/support.js:159mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
runtime.adoptParsed(rootName, parsed); if (!window.__resources) { fetch(location.href).then((res) => res.ok ? res.text() : "").then((t) => { const raw = t ? parseDcText(t) : null; if (raw?.template) runtime.updateHtml(rootName, raw.template); - app/handoffs/averray-mobile-operator-control/project/support.js:1206mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
const pre = bundledBlob(url); if (pre) return pre.text(); return fetch(url).then((r) => { if (!r.ok) throw new Error("HTTP " + r.status); return r.text(); - app/handoffs/averray-mobile-operator-control/project/support.js:1651mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
const target = typeof pre === "string" && pre ? pre : url; const blob = bundledBlob(target); (blob ? blob.text() : fetch(target).then((res2) => { if (!res2.ok) { console.error( - app/lib/api/client.ts:94mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
: method === "GET" && mode === "bounded" ? await fetchAppReadWithRetry(url, { ...init, headers }) : await fetch(url, { ...init, headers }); const bodyText = await response.text(); let body: unknown = bodyText; - app/lib/auth/siwe.ts:70mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
if (token) { try { await fetch(apiUrl("/auth/logout"), { method: "POST", headers: { authorization: `Bearer ${token}` }, - app/lib/auth/siwe.ts:100mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
let response: Response; try { response = await fetch(apiUrl("/auth/refresh"), { method: "POST", headers: { authorization: `Bearer ${token}` }, - mcp-server/src/demo/mcp-dual-era-local.js:152mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
async function post({ body, headers = {} }) { const response = await fetch(endpoint, { method: "POST", headers: { - mcp-server/src/demo/mcp-prober-compose-smoke.js:43mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
async function waitHealthy(url) { for (let attempt = 0; attempt < 60; attempt += 1) { try { if ((await fetch(url)).ok) return; } catch {} await new Promise((resolve) => setTimeout(resolve, 250)); } - mcp-server/src/services/pimlico-client.js:140mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
async rpc(url, method, params = []) { const response = await fetch(url, { method: "POST", headers: { - scripts/mcp/polkadot-docs-bridge.mjs:112mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
} const response = await fetch(endpoint, { method: 'POST', headers, - scripts/mcp/polkadot-docs-bridge.mjs:225mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
try { await fetch(endpoint, { method: 'DELETE', headers: { - scripts/ops/check-x402-ramp-readiness.mjs:200mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
requestPath: url.pathname }); const response = await fetch(url, { method: "POST", headers: { - scripts/ops/exercise-async-xcm-request.mjs:213mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
async function readJson(url, bearerToken) { const response = await fetch(url, { headers: { accept: "application/json", - scripts/ops/exercise-async-xcm-request.mjs:226mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
async function postJson(url, body, bearerToken) { const response = await fetch(url, { method: "POST", headers: { - scripts/ops/pool-venue-ceremony.mjs:632mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
const timeout = setTimeout(() => controller.abort(), 10_000); try { const response = await fetch(url, { signal: controller.signal }); if (!response.ok) throw new Error(`HTTP ${response.status}`); return await response.json(); - scripts/ops/pool-venue-dispatch.mjs:823mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
const timeout = setTimeout(() => controller.abort(), 10_000); try { const response = await fetch(url, { signal: controller.signal }); if (!response.ok) throw new Error(`HTTP ${response.status}`); return response.json(); - scripts/ops/review-github-pr-submissions.mjs:26mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
export async function review(options, { request, print = console.log }) { if (options.mode === "list") { print(JSON.stringify(await request("GET", "/admin/verifier/pending"), null, 2)); return 0; } if (!["preview", "settle"].includes(options.mode)) throw new Error(`Unknown review mode: ${options.mode}`); const payload = { sessionId: options.sessionId }; - scripts/ops/review-github-pr-submissions.mjs:29mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
if (!["preview", "settle"].includes(options.mode)) throw new Error(`Unknown review mode: ${options.mode}`); const payload = { sessionId: options.sessionId }; const verdict = await request("POST", "/admin/verifier/run", { ...payload, preview: true }); print(JSON.stringify(verdict, null, 2)); if (options.mode === "preview") return 0; - scripts/ops/review-github-pr-submissions.mjs:33mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
if (options.mode === "preview") return 0; if (verdict.outcome !== options.expect) return 2; print(JSON.stringify(await request("POST", "/admin/verifier/run", { ...payload, expectOutcome: options.expect }), null, 2)); return 0; } - scripts/ops/run-bounty-worker.mjs:733mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
async publish(packet) { const repository = packet.repository.nameWithOwner; const installation = await request("/installation/repositories?per_page=100", { errorCode: "BOUNTY_GITHUB_SCOPE_REFUSED", exitCode: BOUNTY_WORKER_EXIT.githubScopeRefused, - scripts/ops/run-bounty-worker.mjs:750mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
const materialized = await materializeCommitImpl(packet, { runCommandImpl }); try { const base = await request( `/repos/${repository}/git/ref/heads/${encodeURIComponent(packet.repository.baseRef)}`, { - scripts/ops/run-bounty-worker.mjs:767mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
const headBranch = deriveHeadBranch(packet); const existingHead = await request( `/repos/${repository}/git/ref/heads/${encodeURIComponent(headBranch)}`, { allow404: true }, - scripts/ops/run-bounty-worker.mjs:789mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
const owner = repository.split("/")[0]; const pullRequests = await request( `/repos/${repository}/pulls?state=all&head=${encodeURIComponent(`${owner}:${headBranch}`)}`, ); - scripts/ops/run-bounty-worker.mjs:812mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
} const created = await request(`/repos/${repository}/pulls`, { method: "POST", body: pullRequestPayload(packet, headBranch), - scripts/ops/scratch-dogfood-credit.mjs:39mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
if (!args.input) throw new Error("--input is required as JSON, for example '{\"direction\":\"borrow\",\"pledgeShares\":\"10000000\",\"amount\":\"8000000\"}'"); const headers = { authorization: `Bearer ${args.token}`, "content-type": "application/json" }; const infoResponse = await fetch(new URL("/credit", args.baseUrl), { headers }); const info = await infoResponse.json(); if (!infoResponse.ok || info.available !== true) throw new Error(`getCreditInfo refused: ${JSON.stringify(info)}`); - scripts/ops/scratch-dogfood-credit.mjs:42mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
const info = await infoResponse.json(); if (!infoResponse.ok || info.available !== true) throw new Error(`getCreditInfo refused: ${JSON.stringify(info)}`); const buildResponse = await fetch(new URL("/credit/transactions", args.baseUrl), { method: "POST", headers, body: JSON.stringify(args.input) }); const built = await buildResponse.json(); if (!buildResponse.ok || built.available !== true) throw new Error(`buildCreditTransactions refused: ${JSON.stringify(built)}`); - scripts/ops/validate-subscan-xcm-source.mjs:215mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
async function fetchJson(url, init) { const response = await fetch(url, init); if (!response.ok) { throw new Error(`Request failed with HTTP ${response.status} for ${typeof url === "string" ? url : url.toString()}`); - scripts/post_job_bundle.mjs:123mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
async function readJson(url) { const response = await fetch(url, { headers: { accept: "application/json" - scripts/post_job_bundle.mjs:135mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
async function postJson(url, body, bearerToken) { const response = await fetch(url, { method: "POST", headers: { - witness/src/artifacts.mjs:57mcp-js-ssrf-nonliteral-url (medium/low)
An outbound request is made to a URL that is not a literal. If any part is caller-supplied the server can be aimed at internal hosts it can reach and the caller cannot.
let response; try { response = await fetch(artifact.locator.url, { redirect: "follow" }); } catch (error) { throw new ArtifactAcquisitionError(`artifact fetch failed: ${error.message}`, { - app/lib/api/badge-list-pagination.test.mjs:14mcp-js-dynamic-code-execution (critical/medium)
Test code — listed, not deducted: `03` §3 grades the code a user runs. Caller-influenced input reaches an evaluator. eval, the Function constructor and vm.runInNewContext all execute their argument; none of them sandbox anything by default, vm least of all despite the name.
function compile(source, modules) { const exports = {}; vm.runInNewContext(ts.transpileModule(source, { compilerOptions: { module: ts.ModuleKind.CommonJS, target: ts.ScriptTarget.ES2022 } }).outputText, { exports, require: (key) => { - app/lib/api/review-errors.test.mjs:9mcp-js-dynamic-code-execution (critical/medium)
Test code — listed, not deducted: `03` §3 grades the code a user runs. Caller-influenced input reaches an evaluator. eval, the Function constructor and vm.runInNewContext all execute their argument; none of them sandbox anything by default, vm least of all despite the name.
const root = new URL("../../", import.meta.url); const exports = {}; vm.runInNewContext(ts.transpileModule(readFileSync(new URL("lib/api/client.ts", root), "utf8"), { compilerOptions: { module: ts.ModuleKind.CommonJS, target: ts.ScriptTarget.ES2022 } }).outputText, { exports, require: () => ({}) }); - app/lib/api/review-errors.test.mjs:29mcp-js-dynamic-code-execution (critical/medium)
Test code — listed, not deducted: `03` §3 grades the code a user runs. Caller-influenced input reaches an evaluator. eval, the Function constructor and vm.runInNewContext all execute their argument; none of them sandbox anything by default, vm least of all despite the name.
[[{ code: "github_pr_review_overdue" }], true], [[{ code: "github_pr_review_read_failures" }, { code: "github_pr_review_overdue" }], true] ]) assert.equal(vm.runInNewContext(expression, { githubReview: { warnings } }), expected); }); - app/lib/auth/fresh-visitor.test.mjs:20mcp-js-dynamic-code-execution (critical/medium)
Test code — listed, not deducted: `03` §3 grades the code a user runs. Caller-influenced input reaches an evaluator. eval, the Function constructor and vm.runInNewContext all execute their argument; none of them sandbox anything by default, vm least of all despite the name.
compilerOptions: { module: ts.ModuleKind.CommonJS, jsx: ts.JsxEmit.ReactJSX, target: ts.ScriptTarget.ES2022 } }).outputText; vm.runInNewContext(code, { exports, require: (name) => modules[name] ?? require(name), console, AbortSignal, setTimeout, clearTimeout, setInterval, clearInterval, ...globals }); return exports; - app/lib/chain/chain-ticker.test.mjs:9mcp-js-dynamic-code-execution (critical/medium)
Test code — listed, not deducted: `03` §3 grades the code a user runs. Caller-influenced input reaches an evaluator. eval, the Function constructor and vm.runInNewContext all execute their argument; none of them sandbox anything by default, vm least of all despite the name.
test("review backlog does not degrade a fresh healthy chain chip", () => { const exports = {}; vm.runInNewContext(ts.transpileModule(readFileSync(new URL("../chain-ticker.ts", import.meta.url), "utf8"), { compilerOptions: { module: ts.ModuleKind.CommonJS, target: ts.ScriptTarget.ES2022 } }).outputText, { exports }); - app/lib/ui/human-verdict-panel.test.mjs:15mcp-js-dynamic-code-execution (critical/medium)
Test code — listed, not deducted: `03` §3 grades the code a user runs. Caller-influenced input reaches an evaluator. eval, the Function constructor and vm.runInNewContext all execute their argument; none of them sandbox anything by default, vm least of all despite the name.
function load(file, modules) { const exports = {}; vm.runInNewContext(ts.transpileModule(read(file), { compilerOptions: { module: ts.ModuleKind.CommonJS, jsx: ts.JsxEmit.ReactJSX, target: ts.ScriptTarget.ES2022 } }).outputText, { exports, require: (name) => modules[name] ?? require(name) });
Auth Posture
- authentication_model
auth enforcement present at a call site in the source (inbound token verification) and documented in the README; "refused without credentials" (`03` §4) established by reading the code rather than by connecting
- transport_security
declared endpoint(s) are HTTPS: https://api.averray.com/mcp; scored 80, not 100: `03` §4 reserves 100 for HTTPS *with HSTS*, and an HSTS header can only be seen by connecting, which `04` §9 does not do. This is an unverifiable band, not an observed absence.
- secret_handling
Private Key at .github/workflows/deploy-production.yml:325; JSON Web Token at mcp-server/src/core/redact-provider-error.test.js:28; JSON Web Token at scripts/ops/get-admin-refresh-token.test.mjs:36; …and 1023 more matches
- authorization_granularity
no tool registration was recognised anywhere in the source. An MCP server registers tools by definition, so this is a gap in our detection rather than a server with none, and `03` §4's threshold cannot be applied to it. A tool-list handler IS present, so the tools are built in a shape this counter could not enumerate.
Dependency Health
Showing the 50 most severe of 122 findings on this axis. The score was computed over all of them — findings beyond the fourth of a severity still deduct at a tenth of their weight — so the remainder is omitted from this page, not from the arithmetic.
- package-lock.jsonastro 7.1.0 — GHSA-26w7-cxv4-gfx2
direct, CVSS 9.8, fixed in 7.2.8
https://osv.dev/vulnerability/GHSA-26w7-cxv4-gfx2 - package-lock.jsonnext 16.3.0 — GHSA-2xp9-vwfh-vxw4
direct, CVSS 9.5, fixed in 16.3.3
https://osv.dev/vulnerability/GHSA-2xp9-vwfh-vxw4 - package-lock.jsonnext 16.3.0 — GHSA-vcvr-r3jv-pc5j
direct, CVSS 9.5, fixed in 16.3.6
https://osv.dev/vulnerability/GHSA-vcvr-r3jv-pc5j - package-lock.jsonnext 16.3.0 — GHSA-p293-qw3h-jr36
direct, CVSS 9.0, fixed in 16.3.3
https://osv.dev/vulnerability/GHSA-p293-qw3h-jr36 - package-lock.jsonproxy-addr 2.0.7 — GHSA-jqcg-44mw-7w3h
transitive, CVSS 9.1, fixed in 2.0.8
https://osv.dev/vulnerability/GHSA-jqcg-44mw-7w3h - package-lock.jsonnext 16.3.0 — GHSA-cjq9-62q9-8jv4
direct, CVSS 8.3, fixed in 16.3.8
https://osv.dev/vulnerability/GHSA-cjq9-62q9-8jv4 - package-lock.jsonjs-yaml 4.3.1 — GHSA-2883-xcg3-v3hh
direct, CVSS 7.5, fixed in 4.3.2
https://osv.dev/vulnerability/GHSA-2883-xcg3-v3hh - indexer/package-lock.jsonws 8.20.1 — GHSA-96hv-2xvq-fx4p
direct, CVSS 7.5, fixed in 8.21.0
https://osv.dev/vulnerability/GHSA-96hv-2xvq-fx4p - package-lock.jsonws 8.18.0 — GHSA-96hv-2xvq-fx4p
direct, CVSS 7.5, fixed in 8.21.0
https://osv.dev/vulnerability/GHSA-96hv-2xvq-fx4p - package-lock.jsonsharp 0.35.3 — GHSA-rgj7-g3m4-5g8c
transitive, CVSS 8.9, fixed in 0.35.4
https://osv.dev/vulnerability/GHSA-rgj7-g3m4-5g8c - package-lock.jsonsharp 0.35.3 — GHSA-wq5f-xc86-pv6w
transitive, CVSS 8.9, fixed in 0.35.5
https://osv.dev/vulnerability/GHSA-wq5f-xc86-pv6w - package-lock.jsonbraces 3.0.3 — GHSA-vfj7-8cjw-p6xm
transitive, CVSS 8.7
https://osv.dev/vulnerability/GHSA-vfj7-8cjw-p6xm - package-lock.jsonhttp-cache-semantics 4.2.0 — GHSA-ch52-4w7c-c8xp
transitive, CVSS 8.7
https://osv.dev/vulnerability/GHSA-ch52-4w7c-c8xp - indexer/package-lock.jsonsource-map-js 1.2.1 — GHSA-68fv-2mgg-jv7q
transitive, CVSS 8.7, fixed in 1.2.2
https://osv.dev/vulnerability/GHSA-68fv-2mgg-jv7q - package-lock.jsonsource-map-js 1.2.1 — GHSA-68fv-2mgg-jv7q
transitive, CVSS 8.7, fixed in 1.2.2
https://osv.dev/vulnerability/GHSA-68fv-2mgg-jv7q - package-lock.jsonaxios 1.18.1 — GHSA-x97p-jq2g-jp4f
transitive, CVSS 8.3, fixed in 1.20.0
https://osv.dev/vulnerability/GHSA-x97p-jq2g-jp4f - indexer/package-lock.json@graphql-tools/utils 10.11.0 — GHSA-7mx3-vvmw-hjmv
transitive, CVSS 8.2, fixed in 12.0.1
https://osv.dev/vulnerability/GHSA-7mx3-vvmw-hjmv - indexer/package-lock.json@graphql-tools/utils 11.2.2 — GHSA-7mx3-vvmw-hjmv
transitive, CVSS 8.2, fixed in 12.0.1
https://osv.dev/vulnerability/GHSA-7mx3-vvmw-hjmv - indexer/package-lock.json@graphql-tools/utils 12.0.0 — GHSA-7mx3-vvmw-hjmv
transitive, CVSS 8.2, fixed in 12.0.1
https://osv.dev/vulnerability/GHSA-7mx3-vvmw-hjmv - package-lock.json@graphql-tools/utils 10.11.0 — GHSA-7mx3-vvmw-hjmv
transitive, CVSS 8.2, fixed in 12.0.1
https://osv.dev/vulnerability/GHSA-7mx3-vvmw-hjmv - package-lock.json@graphql-tools/utils 11.1.0 — GHSA-7mx3-vvmw-hjmv
transitive, CVSS 8.2, fixed in 12.0.1
https://osv.dev/vulnerability/GHSA-7mx3-vvmw-hjmv - package-lock.jsonaxios 1.18.1 — GHSA-542g-h47m-68v8
transitive, CVSS 8.2, fixed in 1.20.0
https://osv.dev/vulnerability/GHSA-542g-h47m-68v8 - package-lock.jsonaxios 1.18.1 — GHSA-c29m-xwm3-cm6r
transitive, CVSS 8.2, fixed in 1.20.0
https://osv.dev/vulnerability/GHSA-c29m-xwm3-cm6r - package-lock.jsonaxios 1.18.1 — GHSA-mghh-pgcx-3jjj
transitive, CVSS 8.2, fixed in 1.20.0
https://osv.dev/vulnerability/GHSA-mghh-pgcx-3jjj - package-lock.jsondeepmerge-ts 7.1.5 — GHSA-ggr8-5vv4-36mx
transitive, CVSS 8.2, fixed in 8.0.0
https://osv.dev/vulnerability/GHSA-ggr8-5vv4-36mx - package-lock.jsondevalue 5.8.1 — GHSA-mcm9-63f2-9j32
transitive, CVSS 8.2, fixed in 5.9.3
https://osv.dev/vulnerability/GHSA-mcm9-63f2-9j32 - package-lock.jsondevalue 5.8.1 — GHSA-r9w8-h9r3-54w4
transitive, CVSS 8.2, fixed in 5.9.3
https://osv.dev/vulnerability/GHSA-r9w8-h9r3-54w4 - package-lock.jsondevalue 5.8.1 — GHSA-x5rw-q4pp-hg5g
transitive, CVSS 8.2, fixed in 5.9.3
https://osv.dev/vulnerability/GHSA-x5rw-q4pp-hg5g - indexer/package-lock.jsonkysely 0.26.3 — GHSA-wmrf-hv6w-mr66
transitive, CVSS 8.2, fixed in 0.28.12
https://osv.dev/vulnerability/GHSA-wmrf-hv6w-mr66 - package-lock.jsonkysely 0.26.3 — GHSA-wmrf-hv6w-mr66
transitive, CVSS 8.2, fixed in 0.28.12
https://osv.dev/vulnerability/GHSA-wmrf-hv6w-mr66 - package-lock.jsonnanoid 3.3.16 — GHSA-2v37-7h3g-55p8
transitive, CVSS 8.2, fixed in 3.3.18
https://osv.dev/vulnerability/GHSA-2v37-7h3g-55p8 - package-lock.jsonsmol-toml 1.7.0 — GHSA-7w5x-hrqm-74c2
transitive, CVSS 8.2, fixed in 1.7.1
https://osv.dev/vulnerability/GHSA-7w5x-hrqm-74c2 - package-lock.jsonsvgo 4.0.2 — GHSA-w27v-7q3p-w38r
transitive, CVSS 8.2, fixed in 4.1.0
https://osv.dev/vulnerability/GHSA-w27v-7q3p-w38r - indexer/package-lock.jsonvite 5.4.21 — GHSA-fx2h-pf6j-xcff
transitive, CVSS 8.2, fixed in 6.4.3
https://osv.dev/vulnerability/GHSA-fx2h-pf6j-xcff - package-lock.jsonvite 8.0.13 — GHSA-fx2h-pf6j-xcff
transitive, CVSS 8.2, fixed in 8.0.16
https://osv.dev/vulnerability/GHSA-fx2h-pf6j-xcff - indexer/package-lock.jsonkysely 0.26.3 — GHSA-8cpq-38p9-67gx
transitive, CVSS 8.1, fixed in 0.28.14
https://osv.dev/vulnerability/GHSA-8cpq-38p9-67gx - package-lock.jsonkysely 0.26.3 — GHSA-8cpq-38p9-67gx
transitive, CVSS 8.1, fixed in 0.28.14
https://osv.dev/vulnerability/GHSA-8cpq-38p9-67gx - package-lock.jsonaxios 1.18.1 — GHSA-m8m8-qj5v-23w3
transitive, CVSS 7.6, fixed in 1.20.0
https://osv.dev/vulnerability/GHSA-m8m8-qj5v-23w3 - package-lock.json@fastify/busboy 3.2.0 — GHSA-x8mw-p69m-v3mx
transitive, CVSS 7.5, fixed in 3.2.1
https://osv.dev/vulnerability/GHSA-x8mw-p69m-v3mx - package-lock.json@fastify/busboy 3.2.0 — GHSA-xjh9-v7x6-24jw
transitive, CVSS 7.5, fixed in 3.2.1
https://osv.dev/vulnerability/GHSA-xjh9-v7x6-24jw - indexer/package-lock.json@hono/node-server 1.19.5 — GHSA-wc8c-qw6v-h7f6
transitive, CVSS 7.5, fixed in 1.19.10
https://osv.dev/vulnerability/GHSA-wc8c-qw6v-h7f6 - indexer/package-lock.jsonbrace-expansion 2.1.4 — GHSA-6j4f-fj2g-mc7p
transitive, CVSS 7.5, fixed in 2.1.5
https://osv.dev/vulnerability/GHSA-6j4f-fj2g-mc7p - package-lock.jsonbrace-expansion 1.1.18 — GHSA-6j4f-fj2g-mc7p
transitive, CVSS 7.5, fixed in 1.1.19
https://osv.dev/vulnerability/GHSA-6j4f-fj2g-mc7p - package-lock.jsonbrace-expansion 2.1.4 — GHSA-6j4f-fj2g-mc7p
transitive, CVSS 7.5, fixed in 2.1.5
https://osv.dev/vulnerability/GHSA-6j4f-fj2g-mc7p - package-lock.jsonbrace-expansion 5.0.9 — GHSA-6j4f-fj2g-mc7p
transitive, CVSS 7.5, fixed in 5.0.10
https://osv.dev/vulnerability/GHSA-6j4f-fj2g-mc7p - indexer/package-lock.jsonbrace-expansion 2.1.4 — GHSA-qhr7-859c-m2p7
transitive, CVSS 7.5, fixed in 2.1.6
https://osv.dev/vulnerability/GHSA-qhr7-859c-m2p7 - package-lock.jsonbrace-expansion 1.1.18 — GHSA-qhr7-859c-m2p7
transitive, CVSS 7.5, fixed in 1.1.20
https://osv.dev/vulnerability/GHSA-qhr7-859c-m2p7 - package-lock.jsonbrace-expansion 2.1.4 — GHSA-qhr7-859c-m2p7
transitive, CVSS 7.5, fixed in 2.1.6
https://osv.dev/vulnerability/GHSA-qhr7-859c-m2p7 - package-lock.jsonbrace-expansion 5.0.9 — GHSA-qhr7-859c-m2p7
transitive, CVSS 7.5, fixed in 5.0.11
https://osv.dev/vulnerability/GHSA-qhr7-859c-m2p7 - package-lock.jsondevalue 5.8.1 — GHSA-j22f-vq7h-c4qm
transitive, CVSS 7.5, fixed in 5.9.3
https://osv.dev/vulnerability/GHSA-j22f-vq7h-c4qm
Maintenance
- recency
last commit 2 day(s) before 2026-10-16
- issue_responsiveness
median over 13 issue(s) in the last 6 months: 29.0 day(s) to first maintainer comment
- release_cadence
a release exists but fewer than two fall inside `03` §5's 12-month window, so no median gap can be taken — and §5's remaining bands are all defined on that median. A limit of the methodology, not of this repository.
- bus_factor
this repository's 12-month history is longer than the bounded walk `04` §7 performs, and the authors found so far span more than one of `03` §5's bands. A limit of ours, not a finding about the server.
- repository_signals
`03` §5 scores this by quartile WITHIN the server's category, which needs a distribution across every scanned server and so cannot be computed while scanning one. It is a post-pass over the crawl and that pass does not exist yet.
Transparency
- license
packages/mcp/package.json declares 'UNLICENSED' but no LICENSE file ships with the source
- readme_quality
README.md: installation instructions (+25), configuration documented (+25)
- declared_scopes
specific scope or access documentation is present, but `03` §7's 100 and 60 bands both compare it against scopes INFERRED from static analysis, which is not built yet. Scoring 100 on an empty inference would be a vacuous truth — 'all ∅ inferred scopes are documented'.
- changelog
no CHANGELOG in the fetched source. `03` §7's partial credit for tagged releases carrying release notes is not applied here — that needs the forge API, which this check does not call
- security_contact
no SECURITY.md and no disclosure contact in the README, so a finder has no route to report a vulnerability privately
Contest a finding
Every number here comes from a rule reading published source, and rules get things wrong. If something on this page is incorrect — or correct but mitigated by a check the rule cannot see — open an issue and we answer it in writing. A finding we got wrong is removed and the rule is fixed; one that stands is answered on the record.
Anyone can file: the maintainer, a researcher who disagrees with the classification, or a user who tested the server and could not reproduce it. Name this server and the rule or advisory ID shown beside the finding above. We aim to reply within five business days and to resolve within ten.