diff --git a/CHANGELOG.md b/CHANGELOG.md
index fb8f2bc1..fb7861d6 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -9,6 +9,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Public and private scopes now register DNS records in their correct Route53 hosted zone when using `DNS_TYPE=external_dns`, preventing cross-zone record leakage
 - Add configurable main HTTP port for k8s scopes (default 8080) and HTTP support for additional ports
 - Improve **wait deployment active** failure logging: consolidate repeated `Unhealthy` probe events per pod into a single human-readable line, emit a progress heartbeat every 10% of timeout, and surface a targeted suggested fix based on the probe failure mode (port not open / HTTP non-2xx / probe timeout)
+- Add configurable memory and CPU limits, independent from requests, for k8s scope containers
 - Improve **k8s/diagnose** evidence: every check now emits structured evidence following a documented schema (`summary`, `severity`, `affected`, `details`, `suggested_actions`), failure findings embed the relevant pod log slice (current or previous depending on the failure mode), and a new **Application Logs** category surfaces the user-owned `application` container's log tail directly in the UI
 
 ## [1.11.0] - 2026-04-16
diff --git a/k8s/deployment/build_context b/k8s/deployment/build_context
index 1f357deb..31880dc7 100755
--- a/k8s/deployment/build_context
+++ b/k8s/deployment/build_context
@@ -23,6 +23,20 @@ MIN_REPLICAS=$(echo "$MIN_REPLICAS" | awk '{printf "%d", ($1 == int($1) ? $1 : i
 
 DEPLOYMENT_STATUS=$(echo "$CONTEXT" | jq -r ".deployment.status")
 
+# Fill in *_limit capability fields with the corresponding request value when
+# the limit is missing or explicitly null, then clamp any limit below its
+# request up to the request value. The schema rejects limit < request at save
+# time; this is defense-in-depth so the script can never produce an invalid
+# resources block, regardless of how the context was built.
+normalize_capability_limits() {
+  echo "$1" | jq '
+    .scope.capabilities.cpu_millicores_limit //= .scope.capabilities.cpu_millicores
+    | .scope.capabilities.ram_memory_limit //= .scope.capabilities.ram_memory
+    | .scope.capabilities.cpu_millicores_limit = ([.scope.capabilities.cpu_millicores, .scope.capabilities.cpu_millicores_limit] | max)
+    | .scope.capabilities.ram_memory_limit = ([.scope.capabilities.ram_memory, .scope.capabilities.ram_memory_limit] | max)
+  '
+}
+
 validate_status() {
   local action="$1"
   local status="$2"
@@ -313,6 +327,8 @@ CONTEXT=$(echo "$CONTEXT" | jq \
                 main_http_port: ($main_http_port | tonumber)
           }')
 
+CONTEXT=$(normalize_capability_limits "$CONTEXT")
+
 DEPLOYMENT_ID=$(echo "$CONTEXT" | jq -r '.deployment.id')
 
 OUTPUT_DIR="$SERVICE_PATH/output/$SCOPE_ID-$DEPLOYMENT_ID"
diff --git a/k8s/deployment/templates/deployment.yaml.tpl b/k8s/deployment/templates/deployment.yaml.tpl
index 3552c483..44f58d77 100644
--- a/k8s/deployment/templates/deployment.yaml.tpl
+++ b/k8s/deployment/templates/deployment.yaml.tpl
@@ -312,8 +312,8 @@ spec:
             {{ end }}
           resources:
             limits:
-              cpu: {{ .scope.capabilities.cpu_millicores }}m
-              memory: {{ .scope.capabilities.ram_memory }}Mi
+              cpu: {{ .scope.capabilities.cpu_millicores_limit }}m
+              memory: {{ .scope.capabilities.ram_memory_limit }}Mi
             requests:
               cpu: {{ .scope.capabilities.cpu_millicores }}m
               memory: {{ .scope.capabilities.ram_memory }}Mi
diff --git a/k8s/deployment/tests/build_context.bats b/k8s/deployment/tests/build_context.bats
index 690a8ab4..72d9d020 100644
--- a/k8s/deployment/tests/build_context.bats
+++ b/k8s/deployment/tests/build_context.bats
@@ -946,3 +946,95 @@ set_additional_ports() {
 
   assert_equal "$(echo "$CONTEXT" | jq -c '.scope.capabilities.additional_ports')" "[]"
 }
+
+# =============================================================================
+# Capability limits normalization
+# These tests source the real deployment/build_context and assert on the
+# resulting CONTEXT, exercising the full pipeline. Limits default to their
+# corresponding request value when missing or explicitly null; explicit values
+# pass through.
+# =============================================================================
+
+# Patches CONTEXT.scope.capabilities with the given JSON fragment (merged into
+# the existing capabilities object).
+set_capabilities() {
+  CONTEXT=$(echo "$CONTEXT" | jq --argjson v "$1" '.scope.capabilities = (.scope.capabilities + $v)')
+}
+
+@test "capability limits: cpu limit defaults to cpu_millicores when absent" {
+  setup_full_build_context
+  set_capabilities '{"cpu_millicores":500,"ram_memory":1024,"ram_memory_limit":2048}'
+
+  source "$SCRIPT"
+
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.cpu_millicores_limit')" "500"
+}
+
+@test "capability limits: ram limit defaults to ram_memory when absent" {
+  setup_full_build_context
+  set_capabilities '{"cpu_millicores":500,"cpu_millicores_limit":1000,"ram_memory":1024}'
+
+  source "$SCRIPT"
+
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.ram_memory_limit')" "1024"
+}
+
+@test "capability limits: both limits default to their requests when both absent" {
+  setup_full_build_context
+  set_capabilities '{"cpu_millicores":500,"ram_memory":1024}'
+
+  source "$SCRIPT"
+
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.cpu_millicores_limit')" "500"
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.ram_memory_limit')" "1024"
+}
+
+@test "capability limits: explicit null limits fall back to their requests" {
+  setup_full_build_context
+  set_capabilities '{"cpu_millicores":500,"cpu_millicores_limit":null,"ram_memory":1024,"ram_memory_limit":null}'
+
+  source "$SCRIPT"
+
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.cpu_millicores_limit')" "500"
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.ram_memory_limit')" "1024"
+}
+
+@test "capability limits: explicit non-null limits pass through unchanged" {
+  setup_full_build_context
+  set_capabilities '{"cpu_millicores":500,"cpu_millicores_limit":2000,"ram_memory":1024,"ram_memory_limit":4096}'
+
+  source "$SCRIPT"
+
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.cpu_millicores_limit')" "2000"
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.ram_memory_limit')" "4096"
+}
+
+@test "capability limits: cpu limit below request is clamped up to request" {
+  setup_full_build_context
+  set_capabilities '{"cpu_millicores":500,"cpu_millicores_limit":100,"ram_memory":1024,"ram_memory_limit":2048}'
+
+  source "$SCRIPT"
+
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.cpu_millicores_limit')" "500"
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.ram_memory_limit')" "2048"
+}
+
+@test "capability limits: ram limit below request is clamped up to request" {
+  setup_full_build_context
+  set_capabilities '{"cpu_millicores":500,"cpu_millicores_limit":1000,"ram_memory":1024,"ram_memory_limit":64}'
+
+  source "$SCRIPT"
+
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.cpu_millicores_limit')" "1000"
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.ram_memory_limit')" "1024"
+}
+
+@test "capability limits: both limits below their requests are clamped up" {
+  setup_full_build_context
+  set_capabilities '{"cpu_millicores":500,"cpu_millicores_limit":100,"ram_memory":1024,"ram_memory_limit":64}'
+
+  source "$SCRIPT"
+
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.cpu_millicores_limit')" "500"
+  assert_equal "$(echo "$CONTEXT" | jq -r '.scope.capabilities.ram_memory_limit')" "1024"
+}
diff --git a/k8s/specs/service-spec.json.tpl b/k8s/specs/service-spec.json.tpl
index 656e641d..f2cd6507 100644
--- a/k8s/specs/service-spec.json.tpl
+++ b/k8s/specs/service-spec.json.tpl
@@ -5,11 +5,13 @@
       "type":"object",
       "required":[
          "ram_memory",
+         "ram_memory_limit",
          "visibility",
          "autoscaling",
          "health_check",
          "scaling_type",
          "cpu_millicores",
+         "cpu_millicores_limit",
          "fixed_instances",
          "scheduled_stop",
          "additional_ports",
@@ -44,12 +46,22 @@
                "elements":[
                   {
                      "type":"Category",
-                     "label":"Processor",
+                     "label":"Resources",
                      "elements":[
                         {
                            "type":"Control",
                            "label":"CPU Millicores",
                            "scope":"#/properties/cpu_millicores"
+                        },
+                        {
+                           "type":"Control",
+                           "label":"CPU Millicores Limit",
+                           "scope":"#/properties/cpu_millicores_limit"
+                        },
+                        {
+                           "type":"Control",
+                           "label":"RAM Memory Limit",
+                           "scope":"#/properties/ram_memory_limit"
                         }
                      ]
                   },
@@ -356,6 +368,27 @@
             "default":128,
             "description":"Amount of RAM memory to allocate to the container (in MB)"
          },
+         "ram_memory_limit":{
+            "type":["integer","null"],
+            "oneOf":[
+               {"const":null,  "title":"Same as request"},
+               {"const":64,    "title":"64 MB"},
+               {"const":128,   "title":"128 MB"},
+               {"const":256,   "title":"256 MB"},
+               {"const":512,   "title":"512 MB"},
+               {"const":1024,  "title":"1 GB"},
+               {"const":2048,  "title":"2 GB"},
+               {"const":4096,  "title":"4 GB"},
+               {"const":8192,  "title":"8 GB"},
+               {"const":16384, "title":"16 GB"}
+            ],
+            "title":"RAM Memory Limit",
+            "default":null,
+            "minimum":{
+               "$data":"1/ram_memory"
+            },
+            "description":"Maximum memory the container can use (in MB). Pick 'Same as request' to leave it equal to the request value."
+         },
          "visibility":{
             "type":"string",
             "oneOf":[
@@ -490,6 +523,24 @@
             "minimum":100,
             "description":"Amount of CPU to allocate (in millicores, 1000m = 1 CPU core)"
          },
+         "cpu_millicores_limit":{
+            "type":["integer","null"],
+            "oneOf":[
+               {"const":null, "title":"Same as request"},
+               {"const":100,  "title":"100 m"},
+               {"const":250,  "title":"250 m"},
+               {"const":500,  "title":"500 m"},
+               {"const":1000, "title":"1000 m"},
+               {"const":2000, "title":"2000 m"},
+               {"const":4000, "title":"4000 m"}
+            ],
+            "title":"CPU Millicores Limit",
+            "default":null,
+            "minimum":{
+               "$data":"1/cpu_millicores"
+            },
+            "description":"Maximum CPU the container can use (in millicores). Pick 'Same as request' to leave it equal to the request value."
+         },
          "scheduled_stop":{
             "type":"object",
             "title":"Scheduled Stop",