diff --git a/.env.benchmark.example b/.env.benchmark.example
new file mode 100644
index 000000000..6377e1660
--- /dev/null
+++ b/.env.benchmark.example
@@ -0,0 +1,5 @@
+BENCHMARK_TARGET_URL=http://127.0.0.1:4000
+BENCHMARK_ADMIN_TOKEN=
+BENCHMARK_CONCURRENCY=2
+BENCHMARK_REQUESTS_PER_ENDPOINT=8
+BENCHMARK_OUTPUT_DIR=benchmarks/results
diff --git a/.gitignore b/.gitignore
index 1e3ce10dd..e02a5dc48 100644
--- a/.gitignore
+++ b/.gitignore
@@ -3,5 +3,6 @@ node_modules
 dist
 .env
 .env.*
+!.env.benchmark.example
 coverage
 *.log
diff --git a/apps/api/package.json b/apps/api/package.json
index 25fa0e90e..d32d03beb 100644
--- a/apps/api/package.json
+++ b/apps/api/package.json
@@ -5,7 +5,7 @@
   "scripts": {
     "dev": "node src/server.js",
     "start": "node src/server.js",
-    "test": "node --test src/tests"
+    "test": "node --test src/tests/*.js"
   },
   "dependencies": {
     "cors": "^2.8.5",
diff --git a/benchmarks/README.md b/benchmarks/README.md
new file mode 100644
index 000000000..b5f582152
--- /dev/null
+++ b/benchmarks/README.md
@@ -0,0 +1,38 @@
+# API Benchmarks
+
+This benchmark suite covers every currently mounted API endpoint in `apps/api/src/app.js`, plus `/health`.
+
+## Commands
+
+```sh
+npm run benchmark
+npm run benchmark:smoke
+npm run benchmark:coverage
+```
+
+By default the runner starts the Express app in-process on a random loopback port. To run against an already running local or staging server, set `BENCHMARK_TARGET_URL`.
+
+## Configuration
+
+Copy `.env.benchmark.example` into your shell environment or CI secret configuration and adjust values as needed:
+
+- `BENCHMARK_TARGET_URL`: optional target base URL. If unset, the runner starts the local app.
+- `BENCHMARK_ADMIN_TOKEN`: optional bearer token for protected admin routes. If unset for local runs, the runner creates a short-lived benchmark token.
+- `BENCHMARK_CONCURRENCY`: parallel requests per endpoint.
+- `BENCHMARK_REQUESTS_PER_ENDPOINT`: measured requests per endpoint.
+- `BENCHMARK_OUTPUT_DIR`: output directory for JSON and Markdown reports.
+
+Thresholds live in `benchmarks/thresholds.json`. `npm run benchmark:smoke` is the low-concurrency gate command intended for CI, while local full runs provide broader baseline data.
+
+`npm run benchmark:coverage` checks the benchmark endpoint inventory against the API route mounts in `apps/api/src/app.js`.
+
+## Output
+
+Each run writes:
+
+- `benchmarks/results/full-latest.json`
+- `benchmarks/results/full-latest.md`
+- `benchmarks/results/smoke-latest.json`
+- `benchmarks/results/smoke-latest.md`
+
+Reports include p50, p95, p99 latency, p99 time to first byte, requests per second, peak requests per second, and error rate for each endpoint.
diff --git a/benchmarks/endpoints.mjs b/benchmarks/endpoints.mjs
new file mode 100644
index 000000000..9f0a8ae69
--- /dev/null
+++ b/benchmarks/endpoints.mjs
@@ -0,0 +1,186 @@
+export const benchmarkEndpoints = [
+  {
+    name: "health",
+    method: "GET",
+    path: "/health",
+    expectedStatus: 200
+  },
+  {
+    name: "auth-register",
+    method: "POST",
+    path: "/api/auth/register",
+    expectedStatus: 201,
+    json: ({ iteration }) => ({
+      email: `benchmark-user-${Date.now()}-${iteration}@example.com`,
+      password: "benchmark-pass-123",
+      role: "freelancer"
+    })
+  },
+  {
+    name: "auth-login",
+    method: "POST",
+    path: "/api/auth/login",
+    expectedStatus: 200,
+    json: {
+      email: "benchmark-user@example.com",
+      password: "benchmark-pass-123"
+    }
+  },
+  {
+    name: "auth-oauth-callback",
+    method: "GET",
+    path: "/api/auth/oauth/benchmark/callback",
+    expectedStatus: 200
+  },
+  {
+    name: "auth-refresh",
+    method: "POST",
+    path: "/api/auth/refresh",
+    expectedStatus: 200
+  },
+  {
+    name: "users-list",
+    method: "GET",
+    path: "/api/users",
+    expectedStatus: 200
+  },
+  {
+    name: "users-create",
+    method: "POST",
+    path: "/api/users",
+    expectedStatus: 201,
+    json: ({ iteration }) => ({
+      email: `benchmark-client-${iteration}@example.com`,
+      name: "Benchmark Client",
+      role: "client"
+    })
+  },
+  {
+    name: "jobs-list",
+    method: "GET",
+    path: "/api/jobs",
+    expectedStatus: 200
+  },
+  {
+    name: "jobs-create",
+    method: "POST",
+    path: "/api/jobs",
+    expectedStatus: 201,
+    json: {
+      title: "Benchmark API implementation",
+      description: "Create a benchmark suite with realistic local payloads.",
+      budgetMin: 500,
+      budgetMax: 1500,
+      categoryId: "software-development",
+      skills: ["node", "express", "benchmarking"]
+    }
+  },
+  {
+    name: "proposals-list",
+    method: "GET",
+    path: "/api/proposals",
+    expectedStatus: 200
+  },
+  {
+    name: "proposals-create",
+    method: "POST",
+    path: "/api/proposals",
+    expectedStatus: 201,
+    json: {
+      jobId: "job_benchmark",
+      freelancerId: "usr_benchmark",
+      coverLetter: "I can deliver a reproducible API benchmark suite.",
+      price: 950,
+      timelineDays: 5
+    }
+  },
+  {
+    name: "payments-create",
+    method: "POST",
+    path: "/api/payments",
+    expectedStatus: 201,
+    json: {
+      amount: 95000,
+      currency: "usd",
+      jobId: "job_benchmark"
+    }
+  },
+  {
+    name: "reviews-list",
+    method: "GET",
+    path: "/api/reviews",
+    expectedStatus: 200
+  },
+  {
+    name: "reviews-create",
+    method: "POST",
+    path: "/api/reviews",
+    expectedStatus: 201,
+    json: {
+      jobId: "job_benchmark",
+      reviewerId: "usr_client",
+      revieweeId: "usr_freelancer",
+      rating: 5,
+      comment: "Fast delivery and clear reporting."
+    }
+  },
+  {
+    name: "messages-list",
+    method: "GET",
+    path: "/api/messages",
+    expectedStatus: 200
+  },
+  {
+    name: "messages-create",
+    method: "POST",
+    path: "/api/messages",
+    expectedStatus: 201,
+    json: {
+      fromUserId: "usr_client",
+      toUserId: "usr_freelancer",
+      body: "Can you share the latest benchmark report?"
+    }
+  },
+  {
+    name: "notifications-list",
+    method: "GET",
+    path: "/api/notifications",
+    expectedStatus: 200
+  },
+  {
+    name: "notifications-create",
+    method: "POST",
+    path: "/api/notifications",
+    expectedStatus: 201,
+    json: {
+      userId: "usr_client",
+      type: "proposal_received",
+      message: "A freelancer submitted a proposal."
+    }
+  },
+  {
+    name: "uploads-create",
+    method: "POST",
+    path: "/api/uploads",
+    expectedStatus: 201,
+    multipart: {
+      fieldName: "file",
+      filename: "benchmark-profile.txt",
+      type: "text/plain",
+      body: "Synthetic benchmark upload payload for API route coverage."
+    }
+  },
+  {
+    name: "search",
+    method: "GET",
+    path: "/api/search?q=benchmark",
+    expectedStatus: 200
+  },
+  {
+    name: "admin-metrics",
+    method: "GET",
+    path: "/api/admin/metrics",
+    expectedStatus: 200,
+    auth: true
+  }
+];
diff --git a/benchmarks/results/full-latest.json b/benchmarks/results/full-latest.json
new file mode 100644
index 000000000..5eed1522f
--- /dev/null
+++ b/benchmarks/results/full-latest.json
@@ -0,0 +1,651 @@
+{
+  "mode": "full",
+  "target": "http://127.0.0.1:51582",
+  "generatedAt": "2026-05-20T20:29:40.993Z",
+  "configuration": {
+    "concurrency": 2,
+    "requestsPerEndpoint": 8,
+    "thresholdsFile": "benchmarks/thresholds.json"
+  },
+  "summary": {
+    "endpointCount": 21,
+    "totalRequests": 168,
+    "totalErrors": 0,
+    "errorRate": 0,
+    "maxP99LatencyMs": 13.01,
+    "maxP99TtfbMs": 12.74,
+    "passed": true
+  },
+  "results": [
+    {
+      "name": "health",
+      "method": "GET",
+      "path": "/health",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 390.66,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 4.74,
+        "p50": 1.87,
+        "p95": 13.01,
+        "p99": 13.01
+      },
+      "ttfbMs": {
+        "mean": 4.58,
+        "p50": 1.72,
+        "p95": 12.74,
+        "p99": 12.74
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "auth-register",
+      "method": "POST",
+      "path": "/api/auth/register",
+      "expectedStatus": 201,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 574.19,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 3.34,
+        "p50": 1.37,
+        "p95": 8.73,
+        "p99": 8.73
+      },
+      "ttfbMs": {
+        "mean": 3.29,
+        "p50": 1.34,
+        "p95": 8.55,
+        "p99": 8.55
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "auth-login",
+      "method": "POST",
+      "path": "/api/auth/login",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1491.7,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 1.3,
+        "p50": 1.32,
+        "p95": 1.64,
+        "p99": 1.64
+      },
+      "ttfbMs": {
+        "mean": 1.26,
+        "p50": 1.29,
+        "p95": 1.58,
+        "p99": 1.58
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "auth-oauth-callback",
+      "method": "GET",
+      "path": "/api/auth/oauth/benchmark/callback",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3102.73,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.61,
+        "p50": 0.53,
+        "p95": 0.9,
+        "p99": 0.9
+      },
+      "ttfbMs": {
+        "mean": 0.58,
+        "p50": 0.5,
+        "p95": 0.86,
+        "p99": 0.86
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "auth-refresh",
+      "method": "POST",
+      "path": "/api/auth/refresh",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3310.46,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.59,
+        "p50": 0.55,
+        "p95": 0.75,
+        "p99": 0.75
+      },
+      "ttfbMs": {
+        "mean": 0.56,
+        "p50": 0.53,
+        "p95": 0.72,
+        "p99": 0.72
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "users-list",
+      "method": "GET",
+      "path": "/api/users",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3269.87,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.58,
+        "p50": 0.47,
+        "p95": 0.91,
+        "p99": 0.91
+      },
+      "ttfbMs": {
+        "mean": 0.54,
+        "p50": 0.44,
+        "p95": 0.89,
+        "p99": 0.89
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "users-create",
+      "method": "POST",
+      "path": "/api/users",
+      "expectedStatus": 201,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 2616.98,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.73,
+        "p50": 0.62,
+        "p95": 1.14,
+        "p99": 1.14
+      },
+      "ttfbMs": {
+        "mean": 0.7,
+        "p50": 0.6,
+        "p95": 1.12,
+        "p99": 1.12
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "jobs-list",
+      "method": "GET",
+      "path": "/api/jobs",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3787.51,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.51,
+        "p50": 0.5,
+        "p95": 0.57,
+        "p99": 0.57
+      },
+      "ttfbMs": {
+        "mean": 0.48,
+        "p50": 0.47,
+        "p95": 0.55,
+        "p99": 0.55
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "jobs-create",
+      "method": "POST",
+      "path": "/api/jobs",
+      "expectedStatus": 201,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 2067.54,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.95,
+        "p50": 0.78,
+        "p95": 1.57,
+        "p99": 1.57
+      },
+      "ttfbMs": {
+        "mean": 0.92,
+        "p50": 0.75,
+        "p95": 1.54,
+        "p99": 1.54
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "proposals-list",
+      "method": "GET",
+      "path": "/api/proposals",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 4142.3,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.47,
+        "p50": 0.46,
+        "p95": 0.55,
+        "p99": 0.55
+      },
+      "ttfbMs": {
+        "mean": 0.44,
+        "p50": 0.43,
+        "p95": 0.53,
+        "p99": 0.53
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "proposals-create",
+      "method": "POST",
+      "path": "/api/proposals",
+      "expectedStatus": 201,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3359.05,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.58,
+        "p50": 0.57,
+        "p95": 0.7,
+        "p99": 0.7
+      },
+      "ttfbMs": {
+        "mean": 0.55,
+        "p50": 0.54,
+        "p95": 0.67,
+        "p99": 0.67
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "payments-create",
+      "method": "POST",
+      "path": "/api/payments",
+      "expectedStatus": 201,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3376.24,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.58,
+        "p50": 0.57,
+        "p95": 0.67,
+        "p99": 0.67
+      },
+      "ttfbMs": {
+        "mean": 0.55,
+        "p50": 0.55,
+        "p95": 0.64,
+        "p99": 0.64
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "reviews-list",
+      "method": "GET",
+      "path": "/api/reviews",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3575.88,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.53,
+        "p50": 0.46,
+        "p95": 0.75,
+        "p99": 0.75
+      },
+      "ttfbMs": {
+        "mean": 0.5,
+        "p50": 0.44,
+        "p95": 0.72,
+        "p99": 0.72
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "reviews-create",
+      "method": "POST",
+      "path": "/api/reviews",
+      "expectedStatus": 201,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 2460.75,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.79,
+        "p50": 0.59,
+        "p95": 1.44,
+        "p99": 1.44
+      },
+      "ttfbMs": {
+        "mean": 0.76,
+        "p50": 0.57,
+        "p95": 1.35,
+        "p99": 1.35
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "messages-list",
+      "method": "GET",
+      "path": "/api/messages",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 4278.26,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.45,
+        "p50": 0.45,
+        "p95": 0.54,
+        "p99": 0.54
+      },
+      "ttfbMs": {
+        "mean": 0.43,
+        "p50": 0.42,
+        "p95": 0.51,
+        "p99": 0.51
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "messages-create",
+      "method": "POST",
+      "path": "/api/messages",
+      "expectedStatus": 201,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1984.31,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.99,
+        "p50": 0.77,
+        "p95": 1.73,
+        "p99": 1.73
+      },
+      "ttfbMs": {
+        "mean": 0.96,
+        "p50": 0.74,
+        "p95": 1.7,
+        "p99": 1.7
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "notifications-list",
+      "method": "GET",
+      "path": "/api/notifications",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 4168.84,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.46,
+        "p50": 0.44,
+        "p95": 0.57,
+        "p99": 0.57
+      },
+      "ttfbMs": {
+        "mean": 0.44,
+        "p50": 0.41,
+        "p95": 0.55,
+        "p99": 0.55
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "notifications-create",
+      "method": "POST",
+      "path": "/api/notifications",
+      "expectedStatus": 201,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3563.67,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.54,
+        "p50": 0.53,
+        "p95": 0.66,
+        "p99": 0.66
+      },
+      "ttfbMs": {
+        "mean": 0.52,
+        "p50": 0.51,
+        "p95": 0.64,
+        "p99": 0.64
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "uploads-create",
+      "method": "POST",
+      "path": "/api/uploads",
+      "expectedStatus": 201,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 939.32,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 2.05,
+        "p50": 1.17,
+        "p95": 4.86,
+        "p99": 4.86
+      },
+      "ttfbMs": {
+        "mean": 2.02,
+        "p50": 1.15,
+        "p95": 4.83,
+        "p99": 4.83
+      },
+      "threshold": {
+        "p99LatencyMs": 1500,
+        "p99TtfbMs": 1500,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "search",
+      "method": "GET",
+      "path": "/api/search?q=benchmark",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3003.28,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.65,
+        "p50": 0.53,
+        "p95": 1.05,
+        "p99": 1.05
+      },
+      "ttfbMs": {
+        "mean": 0.61,
+        "p50": 0.5,
+        "p95": 1.03,
+        "p99": 1.03
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "admin-metrics",
+      "method": "GET",
+      "path": "/api/admin/metrics",
+      "expectedStatus": 200,
+      "requests": 8,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 3019.67,
+      "peakRequestsPerSecond": 2000,
+      "latencyMs": {
+        "mean": 0.64,
+        "p50": 0.52,
+        "p95": 1.02,
+        "p99": 1.02
+      },
+      "ttfbMs": {
+        "mean": 0.62,
+        "p50": 0.5,
+        "p95": 1,
+        "p99": 1
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    }
+  ]
+}
diff --git a/benchmarks/results/full-latest.md b/benchmarks/results/full-latest.md
new file mode 100644
index 000000000..5e61b0e53
--- /dev/null
+++ b/benchmarks/results/full-latest.md
@@ -0,0 +1,34 @@
+# API Benchmark Report
+
+- Mode: full
+- Target: http://127.0.0.1:51582
+- Generated: 2026-05-20T20:29:40.993Z
+- Endpoints: 21
+- Requests: 168
+- Error rate: 0.00%
+- Max p99 latency: 13.01 ms
+- Max p99 TTFB: 12.74 ms
+
+| Gate | Method | Path | Requests | Error rate | p50 ms | p95 ms | p99 ms | p99 TTFB ms | RPS |
+| --- | --- | --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
+| pass | GET | /health | 8 | 0.00% | 1.87 | 13.01 | 13.01 | 12.74 | 390.66 |
+| pass | POST | /api/auth/register | 8 | 0.00% | 1.37 | 8.73 | 8.73 | 8.55 | 574.19 |
+| pass | POST | /api/auth/login | 8 | 0.00% | 1.32 | 1.64 | 1.64 | 1.58 | 1491.7 |
+| pass | GET | /api/auth/oauth/benchmark/callback | 8 | 0.00% | 0.53 | 0.9 | 0.9 | 0.86 | 3102.73 |
+| pass | POST | /api/auth/refresh | 8 | 0.00% | 0.55 | 0.75 | 0.75 | 0.72 | 3310.46 |
+| pass | GET | /api/users | 8 | 0.00% | 0.47 | 0.91 | 0.91 | 0.89 | 3269.87 |
+| pass | POST | /api/users | 8 | 0.00% | 0.62 | 1.14 | 1.14 | 1.12 | 2616.98 |
+| pass | GET | /api/jobs | 8 | 0.00% | 0.5 | 0.57 | 0.57 | 0.55 | 3787.51 |
+| pass | POST | /api/jobs | 8 | 0.00% | 0.78 | 1.57 | 1.57 | 1.54 | 2067.54 |
+| pass | GET | /api/proposals | 8 | 0.00% | 0.46 | 0.55 | 0.55 | 0.53 | 4142.3 |
+| pass | POST | /api/proposals | 8 | 0.00% | 0.57 | 0.7 | 0.7 | 0.67 | 3359.05 |
+| pass | POST | /api/payments | 8 | 0.00% | 0.57 | 0.67 | 0.67 | 0.64 | 3376.24 |
+| pass | GET | /api/reviews | 8 | 0.00% | 0.46 | 0.75 | 0.75 | 0.72 | 3575.88 |
+| pass | POST | /api/reviews | 8 | 0.00% | 0.59 | 1.44 | 1.44 | 1.35 | 2460.75 |
+| pass | GET | /api/messages | 8 | 0.00% | 0.45 | 0.54 | 0.54 | 0.51 | 4278.26 |
+| pass | POST | /api/messages | 8 | 0.00% | 0.77 | 1.73 | 1.73 | 1.7 | 1984.31 |
+| pass | GET | /api/notifications | 8 | 0.00% | 0.44 | 0.57 | 0.57 | 0.55 | 4168.84 |
+| pass | POST | /api/notifications | 8 | 0.00% | 0.53 | 0.66 | 0.66 | 0.64 | 3563.67 |
+| pass | POST | /api/uploads | 8 | 0.00% | 1.17 | 4.86 | 4.86 | 4.83 | 939.32 |
+| pass | GET | /api/search?q=benchmark | 8 | 0.00% | 0.53 | 1.05 | 1.05 | 1.03 | 3003.28 |
+| pass | GET | /api/admin/metrics | 8 | 0.00% | 0.52 | 1.02 | 1.02 | 1 | 3019.67 |
diff --git a/benchmarks/results/smoke-latest.json b/benchmarks/results/smoke-latest.json
new file mode 100644
index 000000000..f1b88bd52
--- /dev/null
+++ b/benchmarks/results/smoke-latest.json
@@ -0,0 +1,651 @@
+{
+  "mode": "smoke",
+  "target": "http://127.0.0.1:51574",
+  "generatedAt": "2026-05-20T20:29:37.611Z",
+  "configuration": {
+    "concurrency": 1,
+    "requestsPerEndpoint": 1,
+    "thresholdsFile": "benchmarks/thresholds.json"
+  },
+  "summary": {
+    "endpointCount": 21,
+    "totalRequests": 21,
+    "totalErrors": 0,
+    "errorRate": 0,
+    "maxP99LatencyMs": 16.6,
+    "maxP99TtfbMs": 15.2,
+    "passed": true
+  },
+  "results": [
+    {
+      "name": "health",
+      "method": "GET",
+      "path": "/health",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 60.01,
+      "peakRequestsPerSecond": 60.24,
+      "latencyMs": {
+        "mean": 16.6,
+        "p50": 16.6,
+        "p95": 16.6,
+        "p99": 16.6
+      },
+      "ttfbMs": {
+        "mean": 15.2,
+        "p50": 15.2,
+        "p95": 15.2,
+        "p99": 15.2
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "auth-register",
+      "method": "POST",
+      "path": "/api/auth/register",
+      "expectedStatus": 201,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 128.67,
+      "peakRequestsPerSecond": 129.03,
+      "latencyMs": {
+        "mean": 7.75,
+        "p50": 7.75,
+        "p95": 7.75,
+        "p99": 7.75
+      },
+      "ttfbMs": {
+        "mean": 7.56,
+        "p50": 7.56,
+        "p95": 7.56,
+        "p99": 7.56
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "auth-login",
+      "method": "POST",
+      "path": "/api/auth/login",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 355.91,
+      "peakRequestsPerSecond": 357.14,
+      "latencyMs": {
+        "mean": 2.8,
+        "p50": 2.8,
+        "p95": 2.8,
+        "p99": 2.8
+      },
+      "ttfbMs": {
+        "mean": 2.67,
+        "p50": 2.67,
+        "p95": 2.67,
+        "p99": 2.67
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "auth-oauth-callback",
+      "method": "GET",
+      "path": "/api/auth/oauth/benchmark/callback",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 820.32,
+      "peakRequestsPerSecond": 826.45,
+      "latencyMs": {
+        "mean": 1.21,
+        "p50": 1.21,
+        "p95": 1.21,
+        "p99": 1.21
+      },
+      "ttfbMs": {
+        "mean": 1.14,
+        "p50": 1.14,
+        "p95": 1.14,
+        "p99": 1.14
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "auth-refresh",
+      "method": "POST",
+      "path": "/api/auth/refresh",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 465.13,
+      "peakRequestsPerSecond": 467.29,
+      "latencyMs": {
+        "mean": 2.14,
+        "p50": 2.14,
+        "p95": 2.14,
+        "p99": 2.14
+      },
+      "ttfbMs": {
+        "mean": 1.77,
+        "p50": 1.77,
+        "p95": 1.77,
+        "p99": 1.77
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "users-list",
+      "method": "GET",
+      "path": "/api/users",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 670.37,
+      "peakRequestsPerSecond": 675.68,
+      "latencyMs": {
+        "mean": 1.48,
+        "p50": 1.48,
+        "p95": 1.48,
+        "p99": 1.48
+      },
+      "ttfbMs": {
+        "mean": 1.07,
+        "p50": 1.07,
+        "p95": 1.07,
+        "p99": 1.07
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "users-create",
+      "method": "POST",
+      "path": "/api/users",
+      "expectedStatus": 201,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 381.49,
+      "peakRequestsPerSecond": 384.62,
+      "latencyMs": {
+        "mean": 2.6,
+        "p50": 2.6,
+        "p95": 2.6,
+        "p99": 2.6
+      },
+      "ttfbMs": {
+        "mean": 2.04,
+        "p50": 2.04,
+        "p95": 2.04,
+        "p99": 2.04
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "jobs-list",
+      "method": "GET",
+      "path": "/api/jobs",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 915.61,
+      "peakRequestsPerSecond": 925.93,
+      "latencyMs": {
+        "mean": 1.08,
+        "p50": 1.08,
+        "p95": 1.08,
+        "p99": 1.08
+      },
+      "ttfbMs": {
+        "mean": 0.98,
+        "p50": 0.98,
+        "p95": 0.98,
+        "p99": 0.98
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "jobs-create",
+      "method": "POST",
+      "path": "/api/jobs",
+      "expectedStatus": 201,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 262.83,
+      "peakRequestsPerSecond": 263.16,
+      "latencyMs": {
+        "mean": 3.8,
+        "p50": 3.8,
+        "p95": 3.8,
+        "p99": 3.8
+      },
+      "ttfbMs": {
+        "mean": 3.6,
+        "p50": 3.6,
+        "p95": 3.6,
+        "p99": 3.6
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "proposals-list",
+      "method": "GET",
+      "path": "/api/proposals",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 443.45,
+      "peakRequestsPerSecond": 446.43,
+      "latencyMs": {
+        "mean": 2.24,
+        "p50": 2.24,
+        "p95": 2.24,
+        "p99": 2.24
+      },
+      "ttfbMs": {
+        "mean": 2.17,
+        "p50": 2.17,
+        "p95": 2.17,
+        "p99": 2.17
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "proposals-create",
+      "method": "POST",
+      "path": "/api/proposals",
+      "expectedStatus": 201,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 782.98,
+      "peakRequestsPerSecond": 787.4,
+      "latencyMs": {
+        "mean": 1.27,
+        "p50": 1.27,
+        "p95": 1.27,
+        "p99": 1.27
+      },
+      "ttfbMs": {
+        "mean": 1.22,
+        "p50": 1.22,
+        "p95": 1.22,
+        "p99": 1.22
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "payments-create",
+      "method": "POST",
+      "path": "/api/payments",
+      "expectedStatus": 201,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1000,
+      "peakRequestsPerSecond": 1000,
+      "latencyMs": {
+        "mean": 0.74,
+        "p50": 0.74,
+        "p95": 0.74,
+        "p99": 0.74
+      },
+      "ttfbMs": {
+        "mean": 0.69,
+        "p50": 0.69,
+        "p95": 0.69,
+        "p99": 0.69
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "reviews-list",
+      "method": "GET",
+      "path": "/api/reviews",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1000,
+      "peakRequestsPerSecond": 1000,
+      "latencyMs": {
+        "mean": 0.49,
+        "p50": 0.49,
+        "p95": 0.49,
+        "p99": 0.49
+      },
+      "ttfbMs": {
+        "mean": 0.46,
+        "p50": 0.46,
+        "p95": 0.46,
+        "p99": 0.46
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "reviews-create",
+      "method": "POST",
+      "path": "/api/reviews",
+      "expectedStatus": 201,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1000,
+      "peakRequestsPerSecond": 1000,
+      "latencyMs": {
+        "mean": 0.62,
+        "p50": 0.62,
+        "p95": 0.62,
+        "p99": 0.62
+      },
+      "ttfbMs": {
+        "mean": 0.59,
+        "p50": 0.59,
+        "p95": 0.59,
+        "p99": 0.59
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "messages-list",
+      "method": "GET",
+      "path": "/api/messages",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1000,
+      "peakRequestsPerSecond": 1000,
+      "latencyMs": {
+        "mean": 0.32,
+        "p50": 0.32,
+        "p95": 0.32,
+        "p99": 0.32
+      },
+      "ttfbMs": {
+        "mean": 0.29,
+        "p50": 0.29,
+        "p95": 0.29,
+        "p99": 0.29
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "messages-create",
+      "method": "POST",
+      "path": "/api/messages",
+      "expectedStatus": 201,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1000,
+      "peakRequestsPerSecond": 1000,
+      "latencyMs": {
+        "mean": 0.71,
+        "p50": 0.71,
+        "p95": 0.71,
+        "p99": 0.71
+      },
+      "ttfbMs": {
+        "mean": 0.67,
+        "p50": 0.67,
+        "p95": 0.67,
+        "p99": 0.67
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "notifications-list",
+      "method": "GET",
+      "path": "/api/notifications",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1000,
+      "peakRequestsPerSecond": 1000,
+      "latencyMs": {
+        "mean": 0.41,
+        "p50": 0.41,
+        "p95": 0.41,
+        "p99": 0.41
+      },
+      "ttfbMs": {
+        "mean": 0.38,
+        "p50": 0.38,
+        "p95": 0.38,
+        "p99": 0.38
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "notifications-create",
+      "method": "POST",
+      "path": "/api/notifications",
+      "expectedStatus": 201,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 979.19,
+      "peakRequestsPerSecond": 980.39,
+      "latencyMs": {
+        "mean": 1.02,
+        "p50": 1.02,
+        "p95": 1.02,
+        "p99": 1.02
+      },
+      "ttfbMs": {
+        "mean": 0.98,
+        "p50": 0.98,
+        "p95": 0.98,
+        "p99": 0.98
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "uploads-create",
+      "method": "POST",
+      "path": "/api/uploads",
+      "expectedStatus": 201,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 188.74,
+      "peakRequestsPerSecond": 189.04,
+      "latencyMs": {
+        "mean": 5.29,
+        "p50": 5.29,
+        "p95": 5.29,
+        "p99": 5.29
+      },
+      "ttfbMs": {
+        "mean": 5.24,
+        "p50": 5.24,
+        "p95": 5.24,
+        "p99": 5.24
+      },
+      "threshold": {
+        "p99LatencyMs": 1500,
+        "p99TtfbMs": 1500,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "search",
+      "method": "GET",
+      "path": "/api/search?q=benchmark",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1000,
+      "peakRequestsPerSecond": 1000,
+      "latencyMs": {
+        "mean": 0.87,
+        "p50": 0.87,
+        "p95": 0.87,
+        "p99": 0.87
+      },
+      "ttfbMs": {
+        "mean": 0.83,
+        "p50": 0.83,
+        "p95": 0.83,
+        "p99": 0.83
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    },
+    {
+      "name": "admin-metrics",
+      "method": "GET",
+      "path": "/api/admin/metrics",
+      "expectedStatus": 200,
+      "requests": 1,
+      "errors": 0,
+      "errorRate": 0,
+      "requestsPerSecond": 1000,
+      "peakRequestsPerSecond": 1000,
+      "latencyMs": {
+        "mean": 0.87,
+        "p50": 0.87,
+        "p95": 0.87,
+        "p99": 0.87
+      },
+      "ttfbMs": {
+        "mean": 0.84,
+        "p50": 0.84,
+        "p95": 0.84,
+        "p99": 0.84
+      },
+      "threshold": {
+        "p99LatencyMs": 1000,
+        "p99TtfbMs": 1000,
+        "errorRate": 0
+      },
+      "passedThresholds": true,
+      "failures": []
+    }
+  ]
+}
diff --git a/benchmarks/results/smoke-latest.md b/benchmarks/results/smoke-latest.md
new file mode 100644
index 000000000..a6d891437
--- /dev/null
+++ b/benchmarks/results/smoke-latest.md
@@ -0,0 +1,34 @@
+# API Benchmark Report
+
+- Mode: smoke
+- Target: http://127.0.0.1:51574
+- Generated: 2026-05-20T20:29:37.611Z
+- Endpoints: 21
+- Requests: 21
+- Error rate: 0.00%
+- Max p99 latency: 16.6 ms
+- Max p99 TTFB: 15.2 ms
+
+| Gate | Method | Path | Requests | Error rate | p50 ms | p95 ms | p99 ms | p99 TTFB ms | RPS |
+| --- | --- | --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
+| pass | GET | /health | 1 | 0.00% | 16.6 | 16.6 | 16.6 | 15.2 | 60.01 |
+| pass | POST | /api/auth/register | 1 | 0.00% | 7.75 | 7.75 | 7.75 | 7.56 | 128.67 |
+| pass | POST | /api/auth/login | 1 | 0.00% | 2.8 | 2.8 | 2.8 | 2.67 | 355.91 |
+| pass | GET | /api/auth/oauth/benchmark/callback | 1 | 0.00% | 1.21 | 1.21 | 1.21 | 1.14 | 820.32 |
+| pass | POST | /api/auth/refresh | 1 | 0.00% | 2.14 | 2.14 | 2.14 | 1.77 | 465.13 |
+| pass | GET | /api/users | 1 | 0.00% | 1.48 | 1.48 | 1.48 | 1.07 | 670.37 |
+| pass | POST | /api/users | 1 | 0.00% | 2.6 | 2.6 | 2.6 | 2.04 | 381.49 |
+| pass | GET | /api/jobs | 1 | 0.00% | 1.08 | 1.08 | 1.08 | 0.98 | 915.61 |
+| pass | POST | /api/jobs | 1 | 0.00% | 3.8 | 3.8 | 3.8 | 3.6 | 262.83 |
+| pass | GET | /api/proposals | 1 | 0.00% | 2.24 | 2.24 | 2.24 | 2.17 | 443.45 |
+| pass | POST | /api/proposals | 1 | 0.00% | 1.27 | 1.27 | 1.27 | 1.22 | 782.98 |
+| pass | POST | /api/payments | 1 | 0.00% | 0.74 | 0.74 | 0.74 | 0.69 | 1000 |
+| pass | GET | /api/reviews | 1 | 0.00% | 0.49 | 0.49 | 0.49 | 0.46 | 1000 |
+| pass | POST | /api/reviews | 1 | 0.00% | 0.62 | 0.62 | 0.62 | 0.59 | 1000 |
+| pass | GET | /api/messages | 1 | 0.00% | 0.32 | 0.32 | 0.32 | 0.29 | 1000 |
+| pass | POST | /api/messages | 1 | 0.00% | 0.71 | 0.71 | 0.71 | 0.67 | 1000 |
+| pass | GET | /api/notifications | 1 | 0.00% | 0.41 | 0.41 | 0.41 | 0.38 | 1000 |
+| pass | POST | /api/notifications | 1 | 0.00% | 1.02 | 1.02 | 1.02 | 0.98 | 979.19 |
+| pass | POST | /api/uploads | 1 | 0.00% | 5.29 | 5.29 | 5.29 | 5.24 | 188.74 |
+| pass | GET | /api/search?q=benchmark | 1 | 0.00% | 0.87 | 0.87 | 0.87 | 0.83 | 1000 |
+| pass | GET | /api/admin/metrics | 1 | 0.00% | 0.87 | 0.87 | 0.87 | 0.84 | 1000 |
diff --git a/benchmarks/run-benchmarks.mjs b/benchmarks/run-benchmarks.mjs
new file mode 100644
index 000000000..f3d5a428a
--- /dev/null
+++ b/benchmarks/run-benchmarks.mjs
@@ -0,0 +1,264 @@
+import { mkdir, readFile, writeFile } from "node:fs/promises";
+import { createServer } from "node:http";
+import { dirname, resolve } from "node:path";
+import { fileURLToPath } from "node:url";
+import { benchmarkEndpoints } from "./endpoints.mjs";
+
+const rootDir = resolve(dirname(fileURLToPath(import.meta.url)), "..");
+const args = new Set(process.argv.slice(2));
+const isSmoke = args.has("--smoke");
+const outputDir = resolve(rootDir, process.env.BENCHMARK_OUTPUT_DIR ?? "benchmarks/results");
+const concurrency = Number(process.env.BENCHMARK_CONCURRENCY ?? (isSmoke ? 1 : 2));
+const requestsPerEndpoint = Number(process.env.BENCHMARK_REQUESTS_PER_ENDPOINT ?? (isSmoke ? 1 : 8));
+
+const thresholds = JSON.parse(
+  await readFile(resolve(rootDir, "benchmarks/thresholds.json"), "utf8")
+);
+
+function percentile(values, rank) {
+  if (values.length === 0) return 0;
+  const sorted = [...values].sort((a, b) => a - b);
+  const index = Math.ceil((rank / 100) * sorted.length) - 1;
+  return Number(sorted[Math.max(0, Math.min(index, sorted.length - 1))].toFixed(2));
+}
+
+function mean(values) {
+  if (values.length === 0) return 0;
+  return Number((values.reduce((sum, value) => sum + value, 0) / values.length).toFixed(2));
+}
+
+function routeKey(endpoint) {
+  return `${endpoint.method} ${endpoint.path.split("?")[0]}`;
+}
+
+function resolveThreshold(endpoint) {
+  return {
+    ...thresholds.defaults,
+    ...(thresholds.routes?.[routeKey(endpoint)] ?? {})
+  };
+}
+
+async function createLocalTarget() {
+  if (process.env.BENCHMARK_TARGET_URL) {
+    return {
+      baseUrl: process.env.BENCHMARK_TARGET_URL.replace(/\/$/, ""),
+      close: async () => {}
+    };
+  }
+
+  process.env.NODE_ENV ??= "benchmark";
+  const [{ createApp }, { signAccessToken }] = await Promise.all([
+    import("../apps/api/src/app.js"),
+    import("../apps/api/src/utils/jwt.js")
+  ]);
+
+  if (!process.env.BENCHMARK_ADMIN_TOKEN) {
+    process.env.BENCHMARK_ADMIN_TOKEN = signAccessToken({
+      sub: "benchmark-admin",
+      role: "admin"
+    });
+  }
+
+  const server = createServer(createApp());
+  await new Promise((resolveListen) => server.listen(0, "127.0.0.1", resolveListen));
+  const { port } = server.address();
+
+  return {
+    baseUrl: `http://127.0.0.1:${port}`,
+    close: () => new Promise((resolveClose, rejectClose) => {
+      server.close((error) => error ? rejectClose(error) : resolveClose());
+    })
+  };
+}
+
+function buildRequest(endpoint, iteration) {
+  const headers = {};
+  const request = {
+    method: endpoint.method,
+    headers
+  };
+
+  if (endpoint.auth) {
+    headers.authorization = `Bearer ${process.env.BENCHMARK_ADMIN_TOKEN ?? ""}`;
+  }
+
+  if (endpoint.json) {
+    headers["content-type"] = "application/json";
+    const payload = typeof endpoint.json === "function"
+      ? endpoint.json({ iteration })
+      : endpoint.json;
+    request.body = JSON.stringify(payload);
+  }
+
+  if (endpoint.multipart) {
+    const form = new FormData();
+    form.append(
+      endpoint.multipart.fieldName,
+      new Blob([endpoint.multipart.body], { type: endpoint.multipart.type }),
+      endpoint.multipart.filename
+    );
+    request.body = form;
+  }
+
+  return request;
+}
+
+async function runOne(baseUrl, endpoint, iteration) {
+  const startedAt = performance.now();
+  let headersAt = startedAt;
+  let status = 0;
+  let ok = false;
+  let error = null;
+
+  try {
+    const response = await fetch(`${baseUrl}${endpoint.path}`, buildRequest(endpoint, iteration));
+    headersAt = performance.now();
+    status = response.status;
+    await response.arrayBuffer();
+    ok = status === endpoint.expectedStatus;
+  } catch (requestError) {
+    error = requestError instanceof Error ? requestError.message : String(requestError);
+  }
+
+  const endedAt = performance.now();
+  return {
+    ok,
+    status,
+    error,
+    latencyMs: Number((endedAt - startedAt).toFixed(2)),
+    ttfbMs: Number((headersAt - startedAt).toFixed(2))
+  };
+}
+
+async function runEndpoint(baseUrl, endpoint) {
+  const measurements = [];
+  let nextIteration = 0;
+  const startedAt = performance.now();
+
+  async function worker() {
+    while (nextIteration < requestsPerEndpoint) {
+      const iteration = nextIteration;
+      nextIteration += 1;
+      measurements.push(await runOne(baseUrl, endpoint, iteration));
+    }
+  }
+
+  await Promise.all(Array.from({ length: Math.min(concurrency, requestsPerEndpoint) }, worker));
+
+  const durationSeconds = Math.max((performance.now() - startedAt) / 1000, 0.001);
+  const latencies = measurements.map((sample) => sample.latencyMs);
+  const ttfbs = measurements.map((sample) => sample.ttfbMs);
+  const errors = measurements.filter((sample) => !sample.ok);
+  const threshold = resolveThreshold(endpoint);
+
+  const result = {
+    name: endpoint.name,
+    method: endpoint.method,
+    path: endpoint.path,
+    expectedStatus: endpoint.expectedStatus,
+    requests: measurements.length,
+    errors: errors.length,
+    errorRate: Number((errors.length / measurements.length).toFixed(4)),
+    requestsPerSecond: Number((measurements.length / durationSeconds).toFixed(2)),
+    peakRequestsPerSecond: Number((concurrency / Math.max(Math.min(...latencies) / 1000, 0.001)).toFixed(2)),
+    latencyMs: {
+      mean: mean(latencies),
+      p50: percentile(latencies, 50),
+      p95: percentile(latencies, 95),
+      p99: percentile(latencies, 99)
+    },
+    ttfbMs: {
+      mean: mean(ttfbs),
+      p50: percentile(ttfbs, 50),
+      p95: percentile(ttfbs, 95),
+      p99: percentile(ttfbs, 99)
+    },
+    threshold,
+    passedThresholds:
+      errors.length / measurements.length <= threshold.errorRate &&
+      percentile(latencies, 99) <= threshold.p99LatencyMs &&
+      percentile(ttfbs, 99) <= threshold.p99TtfbMs,
+    failures: errors.slice(0, 3)
+  };
+
+  return result;
+}
+
+function markdownReport(report) {
+  const rows = report.results.map((result) => [
+    result.passedThresholds ? "pass" : "fail",
+    result.method,
+    result.path,
+    result.requests,
+    `${(result.errorRate * 100).toFixed(2)}%`,
+    result.latencyMs.p50,
+    result.latencyMs.p95,
+    result.latencyMs.p99,
+    result.ttfbMs.p99,
+    result.requestsPerSecond
+  ]);
+
+  return [
+    "# API Benchmark Report",
+    "",
+    `- Mode: ${report.mode}`,
+    `- Target: ${report.target}`,
+    `- Generated: ${report.generatedAt}`,
+    `- Endpoints: ${report.summary.endpointCount}`,
+    `- Requests: ${report.summary.totalRequests}`,
+    `- Error rate: ${(report.summary.errorRate * 100).toFixed(2)}%`,
+    `- Max p99 latency: ${report.summary.maxP99LatencyMs} ms`,
+    `- Max p99 TTFB: ${report.summary.maxP99TtfbMs} ms`,
+    "",
+    "| Gate | Method | Path | Requests | Error rate | p50 ms | p95 ms | p99 ms | p99 TTFB ms | RPS |",
+    "| --- | --- | --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: |",
+    ...rows.map((row) => `| ${row.join(" | ")} |`),
+    ""
+  ].join("\n");
+}
+
+const target = await createLocalTarget();
+
+try {
+  const results = [];
+  for (const endpoint of benchmarkEndpoints) {
+    results.push(await runEndpoint(target.baseUrl, endpoint));
+  }
+
+  const totalRequests = results.reduce((sum, result) => sum + result.requests, 0);
+  const totalErrors = results.reduce((sum, result) => sum + result.errors, 0);
+  const report = {
+    mode: isSmoke ? "smoke" : "full",
+    target: target.baseUrl,
+    generatedAt: new Date().toISOString(),
+    configuration: {
+      concurrency,
+      requestsPerEndpoint,
+      thresholdsFile: "benchmarks/thresholds.json"
+    },
+    summary: {
+      endpointCount: results.length,
+      totalRequests,
+      totalErrors,
+      errorRate: Number((totalErrors / totalRequests).toFixed(4)),
+      maxP99LatencyMs: Math.max(...results.map((result) => result.latencyMs.p99)),
+      maxP99TtfbMs: Math.max(...results.map((result) => result.ttfbMs.p99)),
+      passed: results.every((result) => result.passedThresholds)
+    },
+    results
+  };
+
+  await mkdir(outputDir, { recursive: true });
+  const suffix = isSmoke ? "smoke" : "full";
+  const jsonPath = resolve(outputDir, `${suffix}-latest.json`);
+  const markdownPath = resolve(outputDir, `${suffix}-latest.md`);
+  await writeFile(jsonPath, `${JSON.stringify(report, null, 2)}\n`);
+  await writeFile(markdownPath, markdownReport(report));
+
+  console.log(markdownReport(report));
+  if (!report.summary.passed) {
+    process.exitCode = 1;
+  }
+} finally {
+  await target.close();
+}
diff --git a/benchmarks/thresholds.json b/benchmarks/thresholds.json
new file mode 100644
index 000000000..846d68f4c
--- /dev/null
+++ b/benchmarks/thresholds.json
@@ -0,0 +1,14 @@
+{
+  "defaults": {
+    "p99LatencyMs": 1000,
+    "p99TtfbMs": 1000,
+    "errorRate": 0
+  },
+  "routes": {
+    "POST /api/uploads": {
+      "p99LatencyMs": 1500,
+      "p99TtfbMs": 1500,
+      "errorRate": 0
+    }
+  }
+}
diff --git a/benchmarks/verify-endpoint-coverage.mjs b/benchmarks/verify-endpoint-coverage.mjs
new file mode 100644
index 000000000..ecafa457e
--- /dev/null
+++ b/benchmarks/verify-endpoint-coverage.mjs
@@ -0,0 +1,33 @@
+import assert from "node:assert/strict";
+import { readFile } from "node:fs/promises";
+import { resolve } from "node:path";
+import { benchmarkEndpoints } from "./endpoints.mjs";
+
+const rootDir = resolve(import.meta.dirname, "..");
+const appSource = await readFile(resolve(rootDir, "apps/api/src/app.js"), "utf8");
+
+const mountedApiPrefixes = [...appSource.matchAll(/app\.use\("([^"]+)"/g)]
+  .map((match) => match[1])
+  .filter((path) => path.startsWith("/api/"));
+
+const endpointPaths = benchmarkEndpoints.map((endpoint) => endpoint.path.split("?")[0]);
+
+for (const prefix of mountedApiPrefixes) {
+  assert(
+    endpointPaths.some((path) => path === prefix || path.startsWith(`${prefix}/`)),
+    `Missing benchmark endpoint for mounted API prefix ${prefix}`
+  );
+}
+
+assert(
+  endpointPaths.includes("/health"),
+  "Missing benchmark endpoint for /health"
+);
+
+const duplicateKeys = benchmarkEndpoints
+  .map((endpoint) => `${endpoint.method} ${endpoint.path}`)
+  .filter((key, index, all) => all.indexOf(key) !== index);
+
+assert.deepEqual(duplicateKeys, [], "Benchmark endpoint list contains duplicate method/path entries");
+
+console.log(`Verified benchmark coverage for ${mountedApiPrefixes.length} API route prefixes and /health.`);
diff --git a/demos/api-benchmark-demo.mp4 b/demos/api-benchmark-demo.mp4
new file mode 100644
index 000000000..04c3de3c1
Binary files /dev/null and b/demos/api-benchmark-demo.mp4 differ
diff --git a/package.json b/package.json
index 675e6e69d..ea26655ab 100644
--- a/package.json
+++ b/package.json
@@ -7,6 +7,9 @@
   ],
   "scripts": {
     "build": "echo \"Run package-specific builds (e.g. npm run build -w apps/web)\"",
+    "benchmark": "node benchmarks/run-benchmarks.mjs",
+    "benchmark:coverage": "node benchmarks/verify-endpoint-coverage.mjs",
+    "benchmark:smoke": "node benchmarks/run-benchmarks.mjs --smoke",
     "lint": "echo \"No root lint configured\"",
     "test": "npm run test -w apps/api"
   }