kapurshr · Nov 12, 2026 · leon1418 · Nov 12, 2026
@@ -31,6 +31,9 @@ clarify.md Step 3 (do not invent new keys/values).
   it paused?" → process_level / filesystem / none.
 - **launch_concurrency**: "At peak, roughly how many new sessions start per second?" → high
   (many) / moderate / low.
+- **expected_scale**: "Over the next few months, roughly how many sessions (conversations or tasks)
+  will the agent handle per day: under 100, 100 to 10,000, or more than 10,000?" → prototype /
+  growing / high; unknown if unsure.
 - **multi_agent**: "One agent, or several working together?" → no / yes.
 - **deployment_preference**: "Would you rather AWS fully manage the agent for you (you just
   describe it — no code to maintain), bring your own agent code, or let me recommend?" →

@@ -25,6 +25,9 @@ Use direct technical terms. Map each answer onto the keys in clarify.md Step 3.
 - **idle_resume**: "On idle-then-resume, do running processes need to continue exactly
   (process_level), is filesystem persistence enough (filesystem), or not needed (none)?"
 - **launch_concurrency**: "Peak new-session launch rate? high (>5/sec) / moderate / low."
+- **expected_scale**: "Expected volume over the next few months? <100 sessions/day (prototype) /
+  100-10K (growing) / >10K (high)." Volume only; shape is `traffic_pattern`, peak rate is
+  `launch_concurrency`.
 - **multi_agent / framework / existing_cluster / multi_cloud / platform_fit**: ask directly.
 - **deployment_preference**: "Do you want a no-code managed agent runtime (AgentCore Harness),
   bring your own framework code (Strands/LangGraph/CrewAI/custom), or either?" Ask early. Only

@@ -129,9 +129,9 @@ single workload), treat it as `agent_session` and proceed.
   question set
   (`session_duration`, `traffic_pattern`, `session_state`, `isolation`,
   `memory_needs`, `multi_agent`, `framework`, `idle_resume`, `compute_tier`,
-  `launch_concurrency`, `instance_type_requirement`).
+  `launch_concurrency`, `instance_type_requirement`, `expected_scale`).
 - **Every other unit: ONE batched delta question** — "How does `<id>` differ from
-  `<primary>`? (session duration / traffic / compute / state / memory / isolation — name
+  `<primary>`? (session duration / traffic / volume / compute / state / memory / isolation — name
   only what differs)". Parse the reply into per-dimension overrides; dimensions the
   user does not mention inherit the primary unit's answers. Non-agent units only need
   the dimensions their workload-classes rules read (traffic, duration, compute) —
@@ -241,6 +241,10 @@ Collect answers for these keys. Legal values are fixed (Plan 1 Data Model):
   wins, routes it to the Instances compute type (capacity provider).
 - `idle_resume`: process_level | filesystem | none | unknown
 - `launch_concurrency`: high | moderate | low | unknown
+- `expected_scale`: prototype | growing | high | unknown. Expected sessions per day over the next
+  few months: prototype under 100, growing 100 to 10,000, high over 10,000. This is volume, not the
+  traffic shape (`traffic_pattern`) or the peak start rate (`launch_concurrency`). Never derive it
+  from `target_maturity`: when it is not asked or answered it stays `unknown`, which scores neutral.
 - `multi_agent`: yes | no | unknown
 - `deployment_preference`: harness | framework | either | unknown — do you want a no-code
   **managed** agent runtime (AgentCore Harness — declare the agent as config, AWS runs the loop),

@@ -20,7 +20,8 @@
     "idle_resume": { "process_level": 1, "filesystem": 4, "none": 2, "unknown": 2 },
     "compute_tier": { "light": 4, "heavy_non_gpu": 3, "gpu": 3, "unknown": 2 },
     "launch_concurrency": { "high": 5, "moderate": 3, "low": 2, "unknown": 2 },
-    "instance_type_requirement": { "yes": 3, "no": 3, "unknown": 2 }
+    "instance_type_requirement": { "yes": 3, "no": 3, "unknown": 2 },
+    "expected_scale": { "prototype": 4, "growing": 3, "high": 3, "unknown": 2 }
   },
   "deployment_models": ["harness", "framework_on_runtime"],
   "compute_types": ["microvms", "instances"],

@@ -20,7 +20,8 @@
     "idle_resume": { "process_level": 1, "filesystem": 1, "none": 2, "unknown": 2 },
     "compute_tier": { "light": 1, "heavy_non_gpu": 4, "gpu": 5, "unknown": 2 },
     "launch_concurrency": { "high": 3, "moderate": 2, "low": 2, "unknown": 2 },
-    "instance_type_requirement": { "yes": 4, "no": 2, "unknown": 2 }
+    "instance_type_requirement": { "yes": 4, "no": 2, "unknown": 2 },
+    "expected_scale": { "prototype": 1, "growing": 2, "high": 4, "unknown": 2 }
   },
   "deployment_models": [],
   "volatile_facts": [{ "key": "regions", "value": "all", "verify_via_mcp": false }]

@@ -20,7 +20,8 @@
     "idle_resume": { "process_level": 1, "filesystem": 1, "none": 2, "unknown": 2 },
     "compute_tier": { "light": 1, "heavy_non_gpu": 4, "gpu": 5, "unknown": 2 },
     "launch_concurrency": { "high": 3, "moderate": 2, "low": 2, "unknown": 2 },
-    "instance_type_requirement": { "yes": 5, "no": 2, "unknown": 2 }
+    "instance_type_requirement": { "yes": 5, "no": 2, "unknown": 2 },
+    "expected_scale": { "prototype": 0, "growing": 1, "high": 4, "unknown": 2 }
   },
   "deployment_models": [],
   "volatile_facts": [{ "key": "regions", "value": "all", "verify_via_mcp": false }]

@@ -44,7 +44,8 @@
     "idle_resume": { "process_level": 0, "filesystem": 1, "none": 3, "unknown": 2 },
     "compute_tier": { "light": 3, "heavy_non_gpu": 1, "gpu": 0, "unknown": 2 },
     "launch_concurrency": { "high": 4, "moderate": 3, "low": 2, "unknown": 2 },
-    "instance_type_requirement": { "yes": 0, "no": 3, "unknown": 2 }
+    "instance_type_requirement": { "yes": 0, "no": 3, "unknown": 2 },
+    "expected_scale": { "prototype": 4, "growing": 3, "high": 2, "unknown": 2 }
   },
   "deployment_models": [],
   "volatile_facts": [{ "key": "timeout", "verification_key": "lambda.timeout", "value": "15m", "verify_via_mcp": true }]

@@ -54,7 +54,8 @@
     "idle_resume": { "process_level": 5, "filesystem": 3, "none": 1, "unknown": 2 },
     "compute_tier": { "light": 2, "heavy_non_gpu": 7, "gpu": 0, "unknown": 2 },
     "launch_concurrency": { "high": 0, "moderate": 2, "low": 3, "unknown": 2 },
-    "instance_type_requirement": { "yes": 0, "no": 3, "unknown": 2 }
+    "instance_type_requirement": { "yes": 0, "no": 3, "unknown": 2 },
+    "expected_scale": { "prototype": 3, "growing": 3, "high": 2, "unknown": 2 }
   },
   "deployment_models": [],
   "volatile_facts": [

@@ -101,6 +101,7 @@
         "compute_tier": { "enum": ["light", "heavy_non_gpu", "gpu", "unknown"] },
         "idle_resume": { "enum": ["process_level", "filesystem", "none", "unknown"] },
         "launch_concurrency": { "enum": ["high", "moderate", "low", "unknown"] },
+        "expected_scale": { "enum": ["prototype", "growing", "high", "unknown"] },
         "multi_agent": { "enum": ["yes", "no", "unknown"] },
         "deployment_preference": { "enum": ["harness", "framework", "either", "unknown"] },
         "framework": { "enum": ["strands", "langgraph", "crewai", "custom", "none", "unknown"] },

@@ -16,7 +16,7 @@
     "session_duration", "traffic_pattern", "platform_fit", "session_state",
     "ops_preference", "isolation", "memory_needs", "multi_agent", "framework",
     "existing_cluster", "multi_cloud", "idle_resume", "compute_tier",
-    "launch_concurrency", "instance_type_requirement",
+    "launch_concurrency", "instance_type_requirement", "expected_scale",
 ]
 
 # Legal answer values per scoring dimension (the closed set the engine reasons about).
@@ -36,6 +36,7 @@
     "compute_tier": ["light", "heavy_non_gpu", "gpu", "unknown"],
     "launch_concurrency": ["high", "moderate", "low", "unknown"],
     "instance_type_requirement": ["yes", "no", "unknown"],
+    "expected_scale": ["prototype", "growing", "high", "unknown"],
 }
 
 DEFAULTS = {

@@ -834,3 +834,67 @@ def test_runtimes_dir_points_at_skill_references():
     parts = RUNTIMES_DIR.parts
     assert parts[-2:] == ("references", "runtimes"), RUNTIMES_DIR
     assert parts[-3] == "agent-advisor", RUNTIMES_DIR
+
+
+# --- B4 (#566 feedback F2): expected_scale -----------------------------------------------------
+
+def test_unknown_scale_scores_neutral_on_every_runtime():
+    # Existing runs and seeds carry no expected_scale; they must score exactly as before.
+    for profile in scoring.load_profiles():
+        assert profile["affinities"]["expected_scale"]["unknown"] == scoring.NEUTRAL_SCORE
+    answers = {"session_duration": "15min_to_8hr", "traffic_pattern": "steady"}
+    assert scoring.score({"answers": answers})["scores"] == scoring.score(
+        {"answers": {**answers, "expected_scale": "unknown"}})["scores"]
+
+
+def _scale_gain(scale):
+    answers = {"session_duration": "15min_to_8hr"}
+    before = scoring.score({"answers": answers})["scores"]
+    after = scoring.score({"answers": {**answers, "expected_scale": scale}})["scores"]
+    return {runtime: after[runtime] - before[runtime] for runtime in before}
+
+
+def test_prototype_scale_favors_scale_to_zero_runtimes():
+    gain = _scale_gain("prototype")
+    assert min(gain["agentcore"], gain["lambda"]) > gain["lambda_microvms"] > gain["ecs"] > gain["eks"]
+
+
+def test_high_scale_favors_container_runtimes():
+    gain = _scale_gain("high")
+    assert min(gain["ecs"], gain["eks"]) > gain["agentcore"] > max(gain["lambda"], gain["lambda_microvms"])
+
+
+_STEADY_AGENT = {"session_duration": "15min_to_8hr", "traffic_pattern": "steady",
+                 "ops_preference": "moderate", "existing_cluster": "none"}
+
+
+def test_prototype_scale_brings_agentcore_into_the_recommendation():
+    answers = {**_STEADY_AGENT, "framework": "strands", "multi_agent": "no", "platform_fit": "none"}
+    assert scoring.score({"answers": answers})["verdict"] == "lambda_microvms"
+    result = scoring.score({"answers": {**answers, "expected_scale": "prototype"}})
+    assert result["co_recommend"] == ["agentcore", "lambda_microvms"]
+
+
+def test_high_scale_brings_ecs_into_the_recommendation():
+    answers = {**_STEADY_AGENT, "framework": "custom", "session_state": "stateful", "idle_resume": "none"}
+    assert scoring.score({"answers": answers})["verdict"] == "lambda_microvms"
+    result = scoring.score({"answers": {**answers, "expected_scale": "high"}})
+    assert result["co_recommend"] == ["ecs", "lambda_microvms"]
+
+
+def test_seed_schema_accepts_expected_scale_and_rejects_other_values():
+    import jsonschema
+    schema = json.loads(
+        (pathlib.Path(scoring.__file__).parent / "schemas" / "seed.json").read_text())
+    jsonschema.validate({"units": {"agent": {"expected_scale": "growing"}}}, schema)
+    with pytest.raises(jsonschema.ValidationError):
+        jsonschema.validate({"units": {"agent": {"expected_scale": "huge"}}}, schema)
+
+
+def test_clarify_documents_every_expected_scale_value():
+    clarify = (pathlib.Path(scoring.__file__).parent.parent / "references" / "phases" / "clarify")
+    legal_line = next(line for line in (clarify / "clarify.md").read_text().splitlines()
+                      if line.startswith("- `expected_scale`:"))
+    assert legal_line.split(":", 1)[1].split(".")[0].strip().split(" | ") == scoring.LEGAL_VALUES["expected_scale"]
+    for audience in ("clarify-business.md", "clarify-technical.md"):
+        assert "**expected_scale**" in (clarify / audience).read_text()