From 4a090ff10eb287377a341ce21e93ab4220a2c58d Mon Sep 17 00:00:00 2001
From: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
Date: Tue, 11 Aug 2026 12:02:56 -0600
Subject: [PATCH 1/5] Rename evaluation span-attribute bag key

Signed-off-by: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
---
 docs/evaluator/experiments.mdx                |  5 ++--
 .../skills/nemo-experiments-upload/SKILL.md   |  8 +++----
 .../references/troubleshooting.md             |  4 ++--
 .../nemo-intake/references/ingest-formats.md  |  9 +++----
 .../record_tau_airline_traces.py              |  4 +++-
 .../experimentalist/experiment_mirror.py      |  2 +-
 .../experimentalist_backend.py                |  6 +++--
 .../nemo-experimentalist/tests/test_otlp.py   | 24 +++++++++----------
 plugins/nemo-insights/testbed/README.md       |  2 +-
 plugins/nemo-insights/testbed/adapters.py     |  2 +-
 plugins/nemo-insights/testbed/otlp_build.py   |  6 +++--
 .../tests/testbed/test_adapters.py            |  2 +-
 .../tests/testbed/test_otlp_build.py          |  4 ++--
 .../tests/testbed/test_reingest.py            |  4 ++--
 .../skills/nemo-experiments-upload/SKILL.md   |  8 +++----
 .../references/troubleshooting.md             |  4 ++--
 .../nemo-intake/references/ingest-formats.md  |  9 +++----
 services/intake/README.md                     |  8 ++++---
 .../scripts/spans/seed_span_type_showcase.py  |  4 ++--
 .../nmp/intake/spans/clickhouse_migrations.py | 24 ++++++++++++++++---
 .../src/nmp/intake/spans/ingest/otlp.py       | 16 ++++++-------
 .../intake/spans/span_attribute_catalog.py    | 12 ++++++----
 .../integration/spans/test_traces_read.py     |  3 +++
 services/intake/tests/test_atif_v17.py        |  4 ++--
 .../tests/test_spans_clickhouse_migrations.py | 10 ++++++--
 services/intake/tests/test_spans_otlp.py      | 24 +++++++++++++++----
 26 files changed, 133 insertions(+), 75 deletions(-)

diff --git a/docs/evaluator/experiments.mdx b/docs/evaluator/experiments.mdx
index 16c3e97649..7a323db086 100644
--- a/docs/evaluator/experiments.mdx
+++ b/docs/evaluator/experiments.mdx
@@ -231,8 +231,9 @@ session with the Evaluation's identity:
 - For Agent Trajectory Interchange Format (ATIF) and chat-completions, add a **top-level**
   `evaluation_context` object to the ingest payload
   carrying `evaluation_id` (the Evaluation's **name**) and `test_case_id`.
-- For OpenTelemetry Protocol (OTLP), set the `nemo.experiment.id` and `nemo.test_case.id` root-span
-  attributes.
+- For OpenTelemetry Protocol (OTLP), set the `nemo.evaluation.name` (the Evaluation's **name**) and
+  `nemo.test_case.id` root-span attributes. The legacy key `nemo.experiment.id` is still accepted for
+  backward compatibility.
 
 The per-evaluator scores on the leaderboard come from **evaluator results** captured on those
 sessions, either automatically from ATIF verifier rewards or explicitly through the evaluator-results
diff --git a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/SKILL.md b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/SKILL.md
index cacd767e42..f1c158157f 100644
--- a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/SKILL.md
+++ b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/SKILL.md
@@ -130,7 +130,7 @@ curl -sf -X POST \
 Pick the ingest endpoint that matches your producer. **Read `../nemo-intake/references/ingest-formats.md` for the full schema and a copy-pasteable example for each.** How you attach evaluation identity depends on the endpoint:
 
 - **ATIF and chat-completions** (JSON body) — add an `evaluation_context = {evaluation_id: "<the Evaluation name>", test_case_id: "<task id>"}` object to the payload.
-- **OTLP** — there is no body field; set `nemo.experiment.id` (the Evaluation **name**) and
+- **OTLP** — there is no body field; set `nemo.evaluation.name` (the Evaluation **name**) and
   `nemo.test_case.id` (the task ID) as **attributes on the root span**. Spans missing these still
   ingest but won't associate to an Evaluation.
 
@@ -176,7 +176,7 @@ You succeeded when `GET .../evaluations/my-eval-baseline` shows:
 - `run_count` ≥ 1 (each ingested session counts as one run), and
 - non-empty `evaluator_names` / `aggregate_scores` if you logged rewards, and/or `cost_usd` if your spans carried cost.
 
-If `run_count` is 0 after ingesting, the traces didn't associate — almost always a wrong evaluation identity: `evaluation_context.evaluation_id` for ATIF/chat-completions, or the `nemo.experiment.id` root-span attribute for OTLP (see Gotchas).
+If `run_count` is 0 after ingesting, the traces didn't associate — almost always a wrong evaluation identity: `evaluation_context.evaluation_id` for ATIF/chat-completions, or the `nemo.evaluation.name` root-span attribute for OTLP (see Gotchas).
 
 ## If verification fails
 
@@ -184,7 +184,7 @@ If `run_count` is 0 after ingesting, the traces didn't associate — almost alwa
 |---|---|---|
 | `400 "…must be created before it can be logged."` | Ingested before the Evaluation existed, or `evaluation_id` doesn't match | Create the Evaluation (step 2); ensure `evaluation_context.evaluation_id` equals its **name** |
 | `422 Unprocessable` on ingest | Unknown/typo'd top-level key (ATIF/chat-completions are `extra="forbid"`) or bad `schema_version` | Check the exact schema in `../nemo-intake/references/ingest-formats.md`; remove stray keys |
-| Ingest 2xx but `run_count` stays 0 | Evaluation identity missing/wrong — `evaluation_context.evaluation_id` (ATIF/chat-completions) or the `nemo.experiment.id` root-span attribute (OTLP) ≠ the Evaluation's name | Attach the identity for your endpoint; use the Evaluation **name**, not its id |
+| Ingest 2xx but `run_count` stays 0 | Evaluation identity missing/wrong — `evaluation_context.evaluation_id` (ATIF/chat-completions) or the `nemo.evaluation.name` root-span attribute (OTLP) ≠ the Evaluation's name | Attach the identity for your endpoint; use the Evaluation **name**, not its id |
 | `503` on GET evaluation / sessions | ClickHouse (telemetry store) not running | Start ClickHouse; rollups and sessions require it |
 | Scores don't show up | Rewards not under `extra.verifier_result.rewards`, or wrong `data_type` on `/evaluator-results` | See `references/troubleshooting.md` |
 
@@ -192,7 +192,7 @@ If `run_count` is 0 after ingesting, the traces didn't associate — almost alwa
 
 - **Create before you log.** The Evaluation entity must exist before any ingest referencing it — otherwise `400`.
 - **`evaluation_id` is the Evaluation's `name`, not its entity id.** But **`experiment_ids` holds the Experiment's `id`.** Different identifiers; easy to swap.
-- **OTLP uses the attribute key `nemo.experiment.id`** (and `nemo.test_case.id`) — the span-attribute key still says "experiment" even though the JSON body field is `evaluation_context`. Set `nemo.experiment.id` on your root span.
+- **OTLP uses the attribute key `nemo.evaluation.name`** (and `nemo.test_case.id`) — set it to the Evaluation's **name** on your root span, matching the `evaluation_id` field the JSON `evaluation_context` carries on the other endpoints. The pre-rename key `nemo.experiment.id` is still accepted on ingest, so older exporters keep associating.
 - **The parent lives at `/experiments`; `/experiment-groups` is a deprecated hidden alias.** Prefer `/experiments`. Evaluations are created and logged under `/evaluations`.
 - **`metadata` is `dict[str, str]`** — stringify non-string values or you'll get a `422`.
 - **ATIF and chat-completions are `extra="forbid"`** (unknown keys → 422); `evaluation_context` itself is lenient (`extra="ignore"`).
diff --git a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/references/troubleshooting.md b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/references/troubleshooting.md
index 8cd5170a5c..7945cf5eb2 100644
--- a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/references/troubleshooting.md
+++ b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/references/troubleshooting.md
@@ -29,7 +29,7 @@ string — read it first.
 
 | Symptom | Cause | Fix |
 |---|---|---|
-| Ingest returned 2xx but `run_count` stays 0 | `evaluation_context` missing, or `evaluation_id` ≠ the Evaluation's name | Attach `evaluation_context`; use the Evaluation **name**. For OTLP, set the span attribute `nemo.experiment.id` on the root span |
+| Ingest returned 2xx but `run_count` stays 0 | `evaluation_context` missing, or `evaluation_id` ≠ the Evaluation's name | Attach `evaluation_context`; use the Evaluation **name**. For OTLP, set the span attribute `nemo.evaluation.name` on the root span (legacy `nemo.experiment.id` still accepted) |
 | No scores on the evaluation | Rewards not under `extra.verifier_result.rewards` (ATIF), or wrong `data_type` (`/evaluator-results`) | ATIF: `extra.verifier_result.rewards = {criterion: value}`. Explicit: `NUMERIC`/`BOOLEAN` need `value`, `CATEGORICAL`/`TEXT` need `string_value` |
 | No cost on the rollup | The producer never emitted cost | Cost is pass-through — set `cost_usd` (chat-completions / ATIF step `metrics`) or `llm.cost.total` / `gen_ai.usage.cost` (OTLP) |
 | `503` on `GET .../evaluations/{name}` or `/sessions` | ClickHouse (telemetry store) not running | Start ClickHouse; rollups, sessions, and metric sorts/filters all need it |
@@ -39,5 +39,5 @@ string — read it first.
 
 - `evaluation_context.evaluation_id` → the Evaluation's **`name`**.
 - `experiment_ids` (on create evaluation) → a list with the Experiment's **`id`**.
-- OTLP evaluation attribute key → **`nemo.experiment.id`** (test case → `nemo.test_case.id`).
+- OTLP evaluation attribute key → **`nemo.evaluation.name`** (legacy `nemo.experiment.id` still accepted; test case → `nemo.test_case.id`).
 - Parent → **`/experiments`** (`/experiment-groups` is a deprecated hidden alias); evaluations → **`/evaluations`**.
diff --git a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-intake/references/ingest-formats.md b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-intake/references/ingest-formats.md
index a029753b8d..e4d5a248b3 100644
--- a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-intake/references/ingest-formats.md
+++ b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-intake/references/ingest-formats.md
@@ -128,11 +128,12 @@ the root span:
 
 | Meaning | Span attribute key |
 |---|---|
-| Evaluation (by name) | **`nemo.experiment.id`** |
+| Evaluation (by name) | **`nemo.evaluation.name`** |
 | Test case | **`nemo.test_case.id`** |
 
-> Note the key is `nemo.experiment.id` (still "experiment"), even though the REST body field elsewhere
-> is `evaluation_context`. Set `nemo.experiment.id` to the Evaluation's **name**.
+> Set `nemo.evaluation.name` to the Evaluation's **name** (not its id), matching the `evaluation_id`
+> field used by the JSON `evaluation_context` on the other endpoints. The pre-rename key
+> `nemo.experiment.id` is still accepted on ingest, so existing exporters keep working.
 
 Cost/token/model attributes are read from standard GenAI / OpenInference keys (first match wins):
 
@@ -158,7 +159,7 @@ export OTEL_EXPORTER_OTLP_TRACES_ENDPOINT="${NMP_BASE_URL}/apis/intake/v2/worksp
 export OTEL_EXPORTER_OTLP_TRACES_PROTOCOL="http/protobuf"
 ```
 
-Then set `nemo.experiment.id` (+ `nemo.test_case.id`) on the root span of each run.
+Then set `nemo.evaluation.name` (+ `nemo.test_case.id`) on the root span of each run.
 
 ---
 
diff --git a/plugins/nemo-experimentalist/examples/tau3-nooa-agent/record_tau_airline_traces.py b/plugins/nemo-experimentalist/examples/tau3-nooa-agent/record_tau_airline_traces.py
index a8590c1907..0ceddda0a5 100644
--- a/plugins/nemo-experimentalist/examples/tau3-nooa-agent/record_tau_airline_traces.py
+++ b/plugins/nemo-experimentalist/examples/tau3-nooa-agent/record_tau_airline_traces.py
@@ -133,6 +133,8 @@ async def _upload_trials(
     trials: list[TrialResult],
     *,
     workspace: str,
+    # TODO: rename experiment_id -> evaluation_name to match the ingested attribute
+    # (nemo.evaluation.name carries the Evaluation's name). Plumb the rename through callers.
     experiment_id: str,
     agent_name: str,
     agent_version: str,
@@ -149,7 +151,7 @@ async def _upload_trials(
             raise RuntimeError(f"Trace {trace_id} was produced by more than one trial")
 
         attrs = {
-            "nemo.experiment.id": experiment_id,
+            "nemo.evaluation.name": experiment_id,
             "nemo.test_case.id": trial.task_id,
             "nemo.trial.id": trial.id,
             "gen_ai.agent.name": agent_name,
diff --git a/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experiment_mirror.py b/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experiment_mirror.py
index 86f1291049..8003f0cb31 100644
--- a/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experiment_mirror.py
+++ b/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experiment_mirror.py
@@ -145,7 +145,7 @@ async def ensure_experiment(self, candidate: Candidate, split: str) -> str:
         name (``opt-<run>-<label>-<split>``).
 
         The name — not the server-assigned id (``experiment-…``) — is what tags the trace's
-        ``nemo.experiment.id``: it is human-readable, greppable, and stable across resumes,
+        ``nemo.evaluation.name``: it is human-readable, greppable, and stable across resumes,
         and matches the Experiment's own ``name`` so the trace still joins back to it.
         """
         gname = group_name(candidate.run_id)
diff --git a/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experimentalist_backend.py b/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experimentalist_backend.py
index 50c4d013e5..957d247eea 100644
--- a/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experimentalist_backend.py
+++ b/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experimentalist_backend.py
@@ -58,6 +58,8 @@ async def _upload_trace_otlp(
     workspace: str,
     ref: ResourceRef,
     *,
+    # TODO: rename experiment_id -> evaluation_name to match the ingested attribute
+    # (nemo.evaluation.name carries the Evaluation's name). Plumb the rename through callers.
     experiment_id: str,
     trial_id: str,
     task_id: str,
@@ -67,7 +69,7 @@ async def _upload_trace_otlp(
 
     path = Path(urlparse(ref.uri).path)
     attrs: dict[str, str] = {
-        "nemo.experiment.id": experiment_id,
+        "nemo.evaluation.name": experiment_id,
         "nemo.test_case.id": task_id,
         "nemo.trial.id": trial_id,
         **(extra_attrs or {}),
@@ -674,7 +676,7 @@ async def _persist_trial(
                 ):
                     rows.append(span.model_dump(mode="json", exclude_none=True))
                 attrs = {
-                    "nemo.experiment.id": experiment_id,
+                    "nemo.evaluation.name": experiment_id,
                     "nemo.test_case.id": trial.task_id,
                     "nemo.trial.id": trial.id,
                     **agent_attrs,
diff --git a/plugins/nemo-experimentalist/tests/test_otlp.py b/plugins/nemo-experimentalist/tests/test_otlp.py
index 754d2e0cbe..d4b14cc5c8 100644
--- a/plugins/nemo-experimentalist/tests/test_otlp.py
+++ b/plugins/nemo-experimentalist/tests/test_otlp.py
@@ -139,9 +139,9 @@ def test_jsonl_parent_span_hex_becomes_bytes(tmp_path):
 
 def test_jsonl_injects_extra_resource_attrs(tmp_path):
     p = _write(tmp_path / "t.jsonl", {"resourceSpans": [{"scopeSpans": [{"spans": [{"traceId": TRACE_HEX}]}]}]})
-    req = _merge(jsonl_to_protobuf(p, {"nemo.experiment.id": "exp-1", "nemo.trial.id": "t1"}))
+    req = _merge(jsonl_to_protobuf(p, {"nemo.evaluation.name": "exp-1", "nemo.trial.id": "t1"}))
     attrs = _resource_attrs(req.resource_spans[0])
-    assert attrs["nemo.experiment.id"].string_value == "exp-1"
+    assert attrs["nemo.evaluation.name"].string_value == "exp-1"
     assert attrs["nemo.trial.id"].string_value == "t1"
 
 
@@ -155,25 +155,25 @@ def test_jsonl_preserves_existing_resource_attrs(tmp_path):
         ]
     }
     p = _write(tmp_path / "t.jsonl", line)
-    attrs = _resource_attrs(_merge(jsonl_to_protobuf(p, {"nemo.experiment.id": "exp-1"})).resource_spans[0])
+    attrs = _resource_attrs(_merge(jsonl_to_protobuf(p, {"nemo.evaluation.name": "exp-1"})).resource_spans[0])
     assert attrs["hostname"].string_value == "h1"  # original kept
-    assert attrs["nemo.experiment.id"].string_value == "exp-1"  # new added
+    assert attrs["nemo.evaluation.name"].string_value == "exp-1"  # new added
 
 
 def test_jsonl_extra_attr_overrides_duplicate_key(tmp_path):
     line = {
         "resourceSpans": [
             {
-                "resource": {"attributes": [{"key": "nemo.experiment.id", "value": {"stringValue": "old"}}]},
+                "resource": {"attributes": [{"key": "nemo.evaluation.name", "value": {"stringValue": "old"}}]},
                 "scopeSpans": [{"spans": [{"traceId": TRACE_HEX}]}],
             }
         ]
     }
     p = _write(tmp_path / "t.jsonl", line)
-    rs = _merge(jsonl_to_protobuf(p, {"nemo.experiment.id": "new"})).resource_spans[0]
+    rs = _merge(jsonl_to_protobuf(p, {"nemo.evaluation.name": "new"})).resource_spans[0]
     keys = [a.key for a in rs.resource.attributes]
-    assert keys.count("nemo.experiment.id") == 1  # not duplicated
-    assert _resource_attrs(rs)["nemo.experiment.id"].string_value == "new"  # overridden
+    assert keys.count("nemo.evaluation.name") == 1  # not duplicated
+    assert _resource_attrs(rs)["nemo.evaluation.name"].string_value == "new"  # overridden
 
 
 def test_jsonl_chunks_split_and_merge_back(tmp_path):
@@ -190,7 +190,7 @@ def test_jsonl_chunks_split_and_merge_back(tmp_path):
 def test_jsonl_empty_file_yields_no_payloads(tmp_path):
     p = tmp_path / "empty.jsonl"
     p.write_text("")
-    assert jsonl_to_protobuf(p, {"nemo.experiment.id": "exp-1"}) == []
+    assert jsonl_to_protobuf(p, {"nemo.evaluation.name": "exp-1"}) == []
 
 
 # ---------------------------------------------------------------------------
@@ -206,9 +206,9 @@ def test_spans_builds_valid_protobuf_with_ids_and_name():
 
 
 def test_spans_injects_resource_attrs():
-    req = _merge(spans_to_protobuf([_span_row()], {"nemo.experiment.id": "exp-42", "nemo.test_case.id": "task1"}))
+    req = _merge(spans_to_protobuf([_span_row()], {"nemo.evaluation.name": "exp-42", "nemo.test_case.id": "task1"}))
     attrs = _resource_attrs(req.resource_spans[0])
-    assert attrs["nemo.experiment.id"].string_value == "exp-42"
+    assert attrs["nemo.evaluation.name"].string_value == "exp-42"
     assert attrs["nemo.test_case.id"].string_value == "task1"
 
 
@@ -286,4 +286,4 @@ def test_spans_each_row_becomes_splittable_resource_spans():
 
 
 def test_spans_empty_rows_yield_no_payloads():
-    assert spans_to_protobuf([], {"nemo.experiment.id": "exp-1"}) == []
+    assert spans_to_protobuf([], {"nemo.evaluation.name": "exp-1"}) == []
diff --git a/plugins/nemo-insights/testbed/README.md b/plugins/nemo-insights/testbed/README.md
index eb18a6beaf..31c2f850ba 100644
--- a/plugins/nemo-insights/testbed/README.md
+++ b/plugins/nemo-insights/testbed/README.md
@@ -183,7 +183,7 @@ Each benchmark `run` reuses **two stable workspaces** per subject — `<workspac
 (the realistic, oracle-free workspace the Analyst evaluates, blind) and
 `<workspace>-oracle` (the answer key + scores, for the UI). The stanza's `workspace`
 is that base name. Runs no longer mint a workspace each. Run isolation comes from
-the per-span `nemo.experiment.id=<run-id>` tag plus the Analyst's `evaluation_id`
+the per-span `nemo.evaluation.name=<run-id>` tag plus the Analyst's `evaluation_id`
 filter (which AND-pins every span read to that run) — that is what scopes the
 analysis. The matching **Experiment** entity registered on the `-oracle` workspace
 is metadata for the UI (run-picker + leaderboard), not the scoping mechanism. So
diff --git a/plugins/nemo-insights/testbed/adapters.py b/plugins/nemo-insights/testbed/adapters.py
index 5b4161ce0c..77a32e47f2 100644
--- a/plugins/nemo-insights/testbed/adapters.py
+++ b/plugins/nemo-insights/testbed/adapters.py
@@ -155,7 +155,7 @@ async def produce(self) -> dict[str, object]:
         domain = str(cfg["domain"])
         base_url = str(cfg["base_url"])
         base = str(cfg["workspace"])  # stable workspace + agent + experiment-group name
-        run_id = mint_agent_id(base)  # the per-run Experiment name + nemo.experiment.id tag
+        run_id = mint_agent_id(base)  # the per-run Experiment name + nemo.evaluation.name tag
         agent = base  # stable agent name across runs
         created_at = datetime.now(timezone.utc).isoformat()
         # Stable workspaces: the realistic (oracle-free, blind-eval) target is always
diff --git a/plugins/nemo-insights/testbed/otlp_build.py b/plugins/nemo-insights/testbed/otlp_build.py
index 7f7f3fe3a1..eb838c476a 100644
--- a/plugins/nemo-insights/testbed/otlp_build.py
+++ b/plugins/nemo-insights/testbed/otlp_build.py
@@ -78,6 +78,8 @@ def sim_to_spans(
     agent_name: str,
     agent_version: str,
     session_id: str,
+    # TODO: rename experiment_id -> evaluation_name to match the ingested attribute
+    # (nemo.evaluation.name carries the Evaluation's name). Plumb the rename through callers.
     experiment_id: str,
     task: dict[str, Any] | None = None,
     include_rewards: bool,
@@ -95,7 +97,7 @@ def sim_to_spans(
     task's ``evaluation_criteria`` only appear when it is set (the realistic twin
     withholds them so the eval is unaided).
 
-    ``experiment_id`` is stamped on every span as ``nemo.experiment.id`` (with
+    ``experiment_id`` is stamped on every span as ``nemo.evaluation.name`` (with
     the sim's task id as ``nemo.test_case.id``) so a run's spans are queryable
     back via the spans filter ``{"evaluation_id": experiment_id}`` — the per-run
     scope that lets many runs share one workspace.
@@ -113,7 +115,7 @@ def _common() -> dict[str, Any]:
         return {
             "gen_ai.conversation.id": session_id,
             "session.id": session_id,
-            "nemo.experiment.id": experiment_id,
+            "nemo.evaluation.name": experiment_id,
             "nemo.test_case.id": test_case_id,
         }
 
diff --git a/plugins/nemo-insights/tests/testbed/test_adapters.py b/plugins/nemo-insights/tests/testbed/test_adapters.py
index ba5fa41157..638732148b 100644
--- a/plugins/nemo-insights/tests/testbed/test_adapters.py
+++ b/plugins/nemo-insights/tests/testbed/test_adapters.py
@@ -332,7 +332,7 @@ async def fake_run_analyst(**kwargs):
     assert len(exported) == 4  # 2 sims x 2 workspaces
     # Every exported span is tagged with the run id.
     for _ws, _sid, spans in exported:
-        assert all(s["attributes"]["nemo.experiment.id"] == record["experiment_id"] for s in spans)
+        assert all(s["attributes"]["nemo.evaluation.name"] == record["experiment_id"] for s in spans)
     assert all("EVALUATOR" not in _kinds(spans) for ws, _, spans in exported if ws == "tau2-airline")
     assert all("EVALUATOR" in _kinds(spans) for ws, _, spans in exported if ws == "tau2-airline-oracle")
     assert len(evals) == 2
diff --git a/plugins/nemo-insights/tests/testbed/test_otlp_build.py b/plugins/nemo-insights/tests/testbed/test_otlp_build.py
index 5e636d91f9..19767cc98e 100644
--- a/plugins/nemo-insights/tests/testbed/test_otlp_build.py
+++ b/plugins/nemo-insights/tests/testbed/test_otlp_build.py
@@ -357,7 +357,7 @@ def test_every_span_carries_experiment_and_test_case_tags():
     )
     assert spans  # at least the AGENT root + EVALUATOR
     for s in spans:
-        assert s["attributes"]["nemo.experiment.id"] == "tau2-airline-20260626-000000-abcd"
+        assert s["attributes"]["nemo.evaluation.name"] == "tau2-airline-20260626-000000-abcd"
         assert s["attributes"]["nemo.test_case.id"] == "7"
 
 
@@ -373,7 +373,7 @@ def test_tags_identical_across_realistic_and_oracle_twins():
     oracle = sim_to_spans(sim, include_rewards=True, **common)
 
     def tag(s):
-        return (s["attributes"]["nemo.experiment.id"], s["attributes"]["nemo.test_case.id"])
+        return (s["attributes"]["nemo.evaluation.name"], s["attributes"]["nemo.test_case.id"])
 
     assert {tag(s) for s in realistic} == {("run-1", "7")}
     assert {tag(s) for s in oracle} == {("run-1", "7")}
diff --git a/plugins/nemo-insights/tests/testbed/test_reingest.py b/plugins/nemo-insights/tests/testbed/test_reingest.py
index 9c3c864671..753a7b2aa8 100644
--- a/plugins/nemo-insights/tests/testbed/test_reingest.py
+++ b/plugins/nemo-insights/tests/testbed/test_reingest.py
@@ -36,7 +36,7 @@ class _StubCatalog:
         _Spec(_Field("model"), ("gen_ai.request.model", "gen_ai.response.model", "llm.model_name")),
         _Spec(_Field("agent_name"), ("gen_ai.agent.name", "llm.agent.name", "agent.name")),
         _Spec(_Field("agent_version"), ("gen_ai.agent.version", "agent.version")),
-        _Spec(_Field("evaluation_id"), ("nemo.experiment.id",)),
+        _Spec(_Field("evaluation_id"), ("nemo.evaluation.name", "nemo.experiment.id")),
         _Spec(_Field("test_case_id"), ("nemo.test_case.id",)),
         _Spec(_Field("input_tokens"), ("gen_ai.usage.input_tokens", "llm.token_count.prompt")),
         _Spec(_Field("prompt_cache_write_tokens"), ("llm.token_count.prompt_details.cache_write",)),
@@ -137,7 +137,7 @@ def test_agent_doc_golden():
         "output.value": "ok",
         # catalog inversion: semantic columns re-emitted under their top-precedence source key
         "gen_ai.agent.name": "smoke-20260626-121437-5559",
-        "nemo.experiment.id": "smoke-20260626-121437-5559-20260626-121438-a833",
+        "nemo.evaluation.name": "smoke-20260626-121437-5559-20260626-121438-a833",
         "nemo.test_case.id": "1",
     }
 
diff --git a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/SKILL.md b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/SKILL.md
index cacd767e42..f1c158157f 100644
--- a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/SKILL.md
+++ b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/SKILL.md
@@ -130,7 +130,7 @@ curl -sf -X POST \
 Pick the ingest endpoint that matches your producer. **Read `../nemo-intake/references/ingest-formats.md` for the full schema and a copy-pasteable example for each.** How you attach evaluation identity depends on the endpoint:
 
 - **ATIF and chat-completions** (JSON body) — add an `evaluation_context = {evaluation_id: "<the Evaluation name>", test_case_id: "<task id>"}` object to the payload.
-- **OTLP** — there is no body field; set `nemo.experiment.id` (the Evaluation **name**) and
+- **OTLP** — there is no body field; set `nemo.evaluation.name` (the Evaluation **name**) and
   `nemo.test_case.id` (the task ID) as **attributes on the root span**. Spans missing these still
   ingest but won't associate to an Evaluation.
 
@@ -176,7 +176,7 @@ You succeeded when `GET .../evaluations/my-eval-baseline` shows:
 - `run_count` ≥ 1 (each ingested session counts as one run), and
 - non-empty `evaluator_names` / `aggregate_scores` if you logged rewards, and/or `cost_usd` if your spans carried cost.
 
-If `run_count` is 0 after ingesting, the traces didn't associate — almost always a wrong evaluation identity: `evaluation_context.evaluation_id` for ATIF/chat-completions, or the `nemo.experiment.id` root-span attribute for OTLP (see Gotchas).
+If `run_count` is 0 after ingesting, the traces didn't associate — almost always a wrong evaluation identity: `evaluation_context.evaluation_id` for ATIF/chat-completions, or the `nemo.evaluation.name` root-span attribute for OTLP (see Gotchas).
 
 ## If verification fails
 
@@ -184,7 +184,7 @@ If `run_count` is 0 after ingesting, the traces didn't associate — almost alwa
 |---|---|---|
 | `400 "…must be created before it can be logged."` | Ingested before the Evaluation existed, or `evaluation_id` doesn't match | Create the Evaluation (step 2); ensure `evaluation_context.evaluation_id` equals its **name** |
 | `422 Unprocessable` on ingest | Unknown/typo'd top-level key (ATIF/chat-completions are `extra="forbid"`) or bad `schema_version` | Check the exact schema in `../nemo-intake/references/ingest-formats.md`; remove stray keys |
-| Ingest 2xx but `run_count` stays 0 | Evaluation identity missing/wrong — `evaluation_context.evaluation_id` (ATIF/chat-completions) or the `nemo.experiment.id` root-span attribute (OTLP) ≠ the Evaluation's name | Attach the identity for your endpoint; use the Evaluation **name**, not its id |
+| Ingest 2xx but `run_count` stays 0 | Evaluation identity missing/wrong — `evaluation_context.evaluation_id` (ATIF/chat-completions) or the `nemo.evaluation.name` root-span attribute (OTLP) ≠ the Evaluation's name | Attach the identity for your endpoint; use the Evaluation **name**, not its id |
 | `503` on GET evaluation / sessions | ClickHouse (telemetry store) not running | Start ClickHouse; rollups and sessions require it |
 | Scores don't show up | Rewards not under `extra.verifier_result.rewards`, or wrong `data_type` on `/evaluator-results` | See `references/troubleshooting.md` |
 
@@ -192,7 +192,7 @@ If `run_count` is 0 after ingesting, the traces didn't associate — almost alwa
 
 - **Create before you log.** The Evaluation entity must exist before any ingest referencing it — otherwise `400`.
 - **`evaluation_id` is the Evaluation's `name`, not its entity id.** But **`experiment_ids` holds the Experiment's `id`.** Different identifiers; easy to swap.
-- **OTLP uses the attribute key `nemo.experiment.id`** (and `nemo.test_case.id`) — the span-attribute key still says "experiment" even though the JSON body field is `evaluation_context`. Set `nemo.experiment.id` on your root span.
+- **OTLP uses the attribute key `nemo.evaluation.name`** (and `nemo.test_case.id`) — set it to the Evaluation's **name** on your root span, matching the `evaluation_id` field the JSON `evaluation_context` carries on the other endpoints. The pre-rename key `nemo.experiment.id` is still accepted on ingest, so older exporters keep associating.
 - **The parent lives at `/experiments`; `/experiment-groups` is a deprecated hidden alias.** Prefer `/experiments`. Evaluations are created and logged under `/evaluations`.
 - **`metadata` is `dict[str, str]`** — stringify non-string values or you'll get a `422`.
 - **ATIF and chat-completions are `extra="forbid"`** (unknown keys → 422); `evaluation_context` itself is lenient (`extra="ignore"`).
diff --git a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/references/troubleshooting.md b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/references/troubleshooting.md
index 8cd5170a5c..7945cf5eb2 100644
--- a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/references/troubleshooting.md
+++ b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/references/troubleshooting.md
@@ -29,7 +29,7 @@ string — read it first.
 
 | Symptom | Cause | Fix |
 |---|---|---|
-| Ingest returned 2xx but `run_count` stays 0 | `evaluation_context` missing, or `evaluation_id` ≠ the Evaluation's name | Attach `evaluation_context`; use the Evaluation **name**. For OTLP, set the span attribute `nemo.experiment.id` on the root span |
+| Ingest returned 2xx but `run_count` stays 0 | `evaluation_context` missing, or `evaluation_id` ≠ the Evaluation's name | Attach `evaluation_context`; use the Evaluation **name**. For OTLP, set the span attribute `nemo.evaluation.name` on the root span (legacy `nemo.experiment.id` still accepted) |
 | No scores on the evaluation | Rewards not under `extra.verifier_result.rewards` (ATIF), or wrong `data_type` (`/evaluator-results`) | ATIF: `extra.verifier_result.rewards = {criterion: value}`. Explicit: `NUMERIC`/`BOOLEAN` need `value`, `CATEGORICAL`/`TEXT` need `string_value` |
 | No cost on the rollup | The producer never emitted cost | Cost is pass-through — set `cost_usd` (chat-completions / ATIF step `metrics`) or `llm.cost.total` / `gen_ai.usage.cost` (OTLP) |
 | `503` on `GET .../evaluations/{name}` or `/sessions` | ClickHouse (telemetry store) not running | Start ClickHouse; rollups, sessions, and metric sorts/filters all need it |
@@ -39,5 +39,5 @@ string — read it first.
 
 - `evaluation_context.evaluation_id` → the Evaluation's **`name`**.
 - `experiment_ids` (on create evaluation) → a list with the Experiment's **`id`**.
-- OTLP evaluation attribute key → **`nemo.experiment.id`** (test case → `nemo.test_case.id`).
+- OTLP evaluation attribute key → **`nemo.evaluation.name`** (legacy `nemo.experiment.id` still accepted; test case → `nemo.test_case.id`).
 - Parent → **`/experiments`** (`/experiment-groups` is a deprecated hidden alias); evaluations → **`/evaluations`**.
diff --git a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-intake/references/ingest-formats.md b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-intake/references/ingest-formats.md
index a029753b8d..e4d5a248b3 100644
--- a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-intake/references/ingest-formats.md
+++ b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-intake/references/ingest-formats.md
@@ -128,11 +128,12 @@ the root span:
 
 | Meaning | Span attribute key |
 |---|---|
-| Evaluation (by name) | **`nemo.experiment.id`** |
+| Evaluation (by name) | **`nemo.evaluation.name`** |
 | Test case | **`nemo.test_case.id`** |
 
-> Note the key is `nemo.experiment.id` (still "experiment"), even though the REST body field elsewhere
-> is `evaluation_context`. Set `nemo.experiment.id` to the Evaluation's **name**.
+> Set `nemo.evaluation.name` to the Evaluation's **name** (not its id), matching the `evaluation_id`
+> field used by the JSON `evaluation_context` on the other endpoints. The pre-rename key
+> `nemo.experiment.id` is still accepted on ingest, so existing exporters keep working.
 
 Cost/token/model attributes are read from standard GenAI / OpenInference keys (first match wins):
 
@@ -158,7 +159,7 @@ export OTEL_EXPORTER_OTLP_TRACES_ENDPOINT="${NMP_BASE_URL}/apis/intake/v2/worksp
 export OTEL_EXPORTER_OTLP_TRACES_PROTOCOL="http/protobuf"
 ```
 
-Then set `nemo.experiment.id` (+ `nemo.test_case.id`) on the root span of each run.
+Then set `nemo.evaluation.name` (+ `nemo.test_case.id`) on the root span of each run.
 
 ---
 
diff --git a/services/intake/README.md b/services/intake/README.md
index 19a50e6668..7b130385eb 100644
--- a/services/intake/README.md
+++ b/services/intake/README.md
@@ -34,7 +34,8 @@ The Experiments feature captures evaluation runs as leaderboard rows. The flow:
 3. Log traces + evaluator results to an ingest endpoint. The Evaluation must exist first. Attach
    evaluation identity per endpoint: **ATIF/Harbor** and **chat-completions** carry it in the JSON body
    as `evaluation_context = {evaluation_id, test_case_id}`; **OTLP** carries it as root-span attributes
-   `nemo.experiment.id` (the Evaluation name) and `nemo.test_case.id`.
+   `nemo.evaluation.name` (the Evaluation name; legacy `nemo.experiment.id` still accepted) and
+   `nemo.test_case.id`.
 4. Read the rollups — `GET /evaluations/{name}` and `.../{name}/sessions` — or view them in Studio
    (behind the `VITE_FF_EXPERIMENT` flag).
 
@@ -43,8 +44,9 @@ mapping, and a troubleshooting table, see the **`nemo-experiments-upload`** agen
 `packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/`.
 
 > Naming note: the entity/API is mid-rename. The API surfaces "Evaluation" and "Experiment Group",
-> but the stored entity is still `Experiment` and the OTLP evaluation attribute key is
-> `nemo.experiment.id`. This doc follows the current in-code naming.
+> but the stored entity is still `Experiment`. The OTLP evaluation attribute key is
+> `nemo.evaluation.name` (the pre-rename `nemo.experiment.id` is still accepted on ingest). This doc
+> follows the current in-code naming.
 
 ## Local Development
 
diff --git a/services/intake/scripts/spans/seed_span_type_showcase.py b/services/intake/scripts/spans/seed_span_type_showcase.py
index 8b2120f845..4dd69692d6 100644
--- a/services/intake/scripts/spans/seed_span_type_showcase.py
+++ b/services/intake/scripts/spans/seed_span_type_showcase.py
@@ -566,10 +566,10 @@ def seed_cancelled_trace(seeder: Seeder) -> None:
 
 
 def seed_eval_context_trace(seeder: Seeder) -> None:
-    """Trace tagged with nemo.experiment.* so evaluation_context/experiment_context populate."""
+    """Trace tagged with the evaluation identity so evaluation_context/experiment_context populate."""
     session = "showcase-eval-run"
     experiment_attrs = {
-        "nemo.experiment.id": "exp-type-showcase",
+        "nemo.evaluation.name": "exp-type-showcase",
         "nemo.experiment.run_id": "run-01",
         "nemo.experiment.sha": "a1b2c3d4",
         "nemo.test_case.id": "case-0007",
diff --git a/services/intake/src/nmp/intake/spans/clickhouse_migrations.py b/services/intake/src/nmp/intake/spans/clickhouse_migrations.py
index 5f6a62cbfa..ef43f01646 100644
--- a/services/intake/src/nmp/intake/spans/clickhouse_migrations.py
+++ b/services/intake/src/nmp/intake/spans/clickhouse_migrations.py
@@ -252,9 +252,21 @@ def _create_trace_index_schema(client, settings: ClickHouseMigrationSettings) ->
     client.command(f"DROP TABLE IF EXISTS {table}")
 
     project_key = spec_for_field(SpanAttributeField.PROJECT).bag_key
-    evaluation_key = spec_for_field(SpanAttributeField.EVALUATION_ID).bag_key
     test_case_key = spec_for_field(SpanAttributeField.TEST_CASE_ID).bag_key
 
+    # Resolve evaluation_id by coalescing the canonical bag key with any legacy aliases. Ingest always
+    # re-keys new spans to the canonical key, so this only matters for the backfill INSERT: spans stored
+    # under an older key (e.g. the pre-rename ``nemo.experiment.id``) keep their evaluation association
+    # instead of being dropped when the MV is rebuilt.
+    evaluation_spec = spec_for_field(SpanAttributeField.EVALUATION_NAME)
+    evaluation_keys = [
+        evaluation_spec.bag_key,
+        *(k for k in evaluation_spec.source_keys if k != evaluation_spec.bag_key),
+    ]
+    evaluation_id_expr = (
+        "coalesce(" + ", ".join(f"nullIf(attributes_string['{key}'], '')" for key in evaluation_keys) + ", '')"
+    )
+
     # Note this is logically a single table. CH requires creating an underlying table and then a view that writes to that table.
     client.command(
         f"""
@@ -310,7 +322,7 @@ def _create_trace_index_schema(client, settings: ClickHouseMigrationSettings) ->
             input AS root_input,
             output AS root_output,
             attributes_string['{project_key}'] AS project,
-            attributes_string['{evaluation_key}'] AS evaluation_id,
+            {evaluation_id_expr} AS evaluation_id,
             attributes_string['{test_case_key}'] AS test_case_id,
             start_time AS root_started_at,
             nullIf(end_time, toDateTime64(0, 6)) AS root_ended_at,
@@ -343,7 +355,7 @@ def _create_trace_index_schema(client, settings: ClickHouseMigrationSettings) ->
     ("ch_evaluator_results_0002", _add_evaluator_results_skip_indexes),
     ("ch_trace_index_0003", _create_trace_index_schema),
     # NeMo-owned span attribute keys moved under the ``nemo.*`` namespace
-    # (e.g., ``evaluation.id`` → ``nemo.evaluation.id``). The trace_index MV's SELECT clause
+    # (e.g., ``test_case.id`` → ``nemo.test_case.id``). The trace_index MV's SELECT clause
     # resolves bag keys from the catalog at creation time, so any environment that already
     # applied 0003 has the old keys baked in. Re-running the schema function drops and
     # recreates the MV with the current catalog keys.
@@ -354,6 +366,12 @@ def _create_trace_index_schema(client, settings: ClickHouseMigrationSettings) ->
     # re-run the rebuild under a new key. The function drops and recreates trace_index with the new
     # column and backfills losslessly from ``spans`` (the durable source of truth).
     ("ch_trace_index_0005_evaluation_id", _create_trace_index_schema),
+    # The evaluation span-attribute bag key was renamed ``nemo.experiment.id`` -> ``nemo.evaluation.name``
+    # (finishing the Experiment->Evaluation rename). The MV bakes the SELECT at creation time, so
+    # environments that applied 0005 still resolve evaluation_id from the old key. Re-run the rebuild:
+    # the MV now coalesces both keys, so spans already ingested under ``nemo.experiment.id`` keep their
+    # evaluation association while new spans use the canonical key.
+    ("ch_trace_index_0006_nemo_evaluation_name", _create_trace_index_schema),
 ]
 CURRENT_SCHEMA_VERSION = _MIGRATIONS[-1][0]
 
diff --git a/services/intake/src/nmp/intake/spans/ingest/otlp.py b/services/intake/src/nmp/intake/spans/ingest/otlp.py
index 9baf83acf5..7c21f5ba09 100644
--- a/services/intake/src/nmp/intake/spans/ingest/otlp.py
+++ b/services/intake/src/nmp/intake/spans/ingest/otlp.py
@@ -24,7 +24,7 @@
 
 # Bag key the evaluation id lands under on a built span (from the attribute catalog). The ingest loop
 # reads it back off each span to learn which evaluations a batch touched, for the denormalizer.
-_EVALUATION_ID_BAG_KEY = spec_for_field(SpanAttributeField.EVALUATION_ID).bag_key
+_EVALUATION_NAME_BAG_KEY = spec_for_field(SpanAttributeField.EVALUATION_NAME).bag_key
 
 # Ordered by precedence. Keep direct input.value/output.value first so existing
 # OpenInference/LangChain payloads continue to win over framework-specific fallbacks.
@@ -84,7 +84,7 @@ async def ingest_otlp_traces(
     export_request = _parse_export_request(body)
     ingested_at = utc_now()
     spans: list[IntakeSpan] = []
-    evaluation_ids: set[str] = set()
+    evaluation_names: set[str] = set()
     errors: list[str] = []
 
     for resource_spans in export_request.resource_spans:
@@ -109,16 +109,16 @@ async def ingest_otlp_traces(
                     errors.append(f"span {span_hex}: {exc}")
                     continue
                 spans.append(span_domain)
-                evaluation_id = span_domain.attributes_string.get(_EVALUATION_ID_BAG_KEY)
-                if evaluation_id:
-                    evaluation_ids.add(evaluation_id)
+                evaluation_name = span_domain.attributes_string.get(_EVALUATION_NAME_BAG_KEY)
+                if evaluation_name:
+                    evaluation_names.add(evaluation_name)
 
     await service.ingest_batch(TraceBatch(spans=spans))
     # A single OTLP batch can carry spans for several evaluations (associated via the
-    # nemo.experiment.id attribute), so refresh the denormalized name facets for every one it touched.
+    # nemo.evaluation.name attribute), so refresh the denormalized name facets for every one it touched.
     if denormalizer is not None:
-        for evaluation_id in evaluation_ids:
-            denormalizer.mark_dirty(workspace=workspace, evaluation_id=evaluation_id)
+        for evaluation_name in evaluation_names:
+            denormalizer.mark_dirty(workspace=workspace, evaluation_id=evaluation_name)
     return IngestResponse(errors=errors)
 
 
diff --git a/services/intake/src/nmp/intake/spans/span_attribute_catalog.py b/services/intake/src/nmp/intake/spans/span_attribute_catalog.py
index 01535bfb94..6ffe05bc92 100644
--- a/services/intake/src/nmp/intake/spans/span_attribute_catalog.py
+++ b/services/intake/src/nmp/intake/spans/span_attribute_catalog.py
@@ -36,7 +36,7 @@ class SpanAttributeField(StrEnum):
     AGENT_VERSION = "agent_version"
     TOOL_NAME = "tool_name"
     PROJECT = "project"
-    EVALUATION_ID = "evaluation_id"
+    EVALUATION_NAME = "evaluation_id"
     TEST_CASE_ID = "test_case_id"
     ERROR_TYPE = "error_type"
     ERROR_MESSAGE = "error_message"
@@ -148,10 +148,14 @@ class AttributeSpec:
         ),
     ),
     AttributeSpec(
-        field=SpanAttributeField.EVALUATION_ID,
+        field=SpanAttributeField.EVALUATION_NAME,
         bag=AttributeBag.STRING,
-        bag_key="nemo.experiment.id",
-        source_keys=("nemo.experiment.id",),
+        # Canonical key is ``nemo.evaluation.name`` (renamed from ``nemo.experiment.id`` in the
+        # Experiment->Evaluation rename). The legacy ``nemo.experiment.id`` is still accepted on ingest
+        # and normalized to the canonical key, so producers can migrate independently without dropping
+        # their evaluation association.
+        bag_key="nemo.evaluation.name",
+        source_keys=("nemo.evaluation.name", "nemo.experiment.id"),
     ),
     AttributeSpec(
         field=SpanAttributeField.TEST_CASE_ID,
diff --git a/services/intake/tests/integration/spans/test_traces_read.py b/services/intake/tests/integration/spans/test_traces_read.py
index dbd8f8d841..b775e5fe25 100644
--- a/services/intake/tests/integration/spans/test_traces_read.py
+++ b/services/intake/tests/integration/spans/test_traces_read.py
@@ -24,6 +24,9 @@ def test_traces_read_returns_core_trace_summary(client: TestClient, make_otlp_re
                     "openinference.span.kind": "AGENT",
                     "gen_ai.conversation.id": "trace-session",
                     "project": "project-a",
+                    # Emit the legacy evaluation key on purpose: ingest normalizes it to the canonical
+                    # nemo.evaluation.name, so this asserts the dual-read path end to end (a pre-rename
+                    # producer still associates its traces to the evaluation).
                     "nemo.experiment.id": "experiment-a",
                     "nemo.test_case.id": "case-a",
                     "deployment.environment.name": "prod",
diff --git a/services/intake/tests/test_atif_v17.py b/services/intake/tests/test_atif_v17.py
index 5a9b09df5c..96e104c908 100644
--- a/services/intake/tests/test_atif_v17.py
+++ b/services/intake/tests/test_atif_v17.py
@@ -510,7 +510,7 @@ def test_atif_mapping_writes_evaluation_context_only_on_root_span() -> None:
 
     root = next(span for span in spans if span.name == "sample-agent")
     child = next(span for span in spans if span.name == "user-1")
-    assert root.attributes_string["nemo.experiment.id"] == EVALUATION_CONTEXT["evaluation_id"]
+    assert root.attributes_string["nemo.evaluation.name"] == EVALUATION_CONTEXT["evaluation_id"]
     # sha/run_id/metadata are dropped by the trimmed ingest EvaluationContext, so they never
     # reach the span from the JSON evaluation_context path.
     assert "nemo.experiment.sha" not in root.attributes_string
@@ -532,7 +532,7 @@ def test_atif_mapping_writes_evaluation_context_only_on_root_span() -> None:
     child_response = Span.from_domain(child)
     assert child_response.evaluation_context is None
     assert "evaluation.id" not in child.attributes_string
-    assert "nemo.experiment.id" not in child.attributes_string
+    assert "nemo.evaluation.name" not in child.attributes_string
     assert "nemo.test_case.id" not in child.attributes_string
 
 
diff --git a/services/intake/tests/test_spans_clickhouse_migrations.py b/services/intake/tests/test_spans_clickhouse_migrations.py
index d50dbc1d2f..4a7847e417 100644
--- a/services/intake/tests/test_spans_clickhouse_migrations.py
+++ b/services/intake/tests/test_spans_clickhouse_migrations.py
@@ -61,7 +61,10 @@ def test_trace_index_schema_is_root_span_projection():
     assert "TO {table}" in ddl
     assert "INSERT INTO {table}" in source
     assert "WHERE external_parent_span_id = ''" in ddl
-    assert "attributes_string['{evaluation_key}'] AS evaluation_id" in ddl
+    # evaluation_id is resolved via a coalesce expression (canonical key + legacy aliases), not a single
+    # bag-key lookup, so the backfill keeps spans stored under an older key associated.
+    assert "{evaluation_id_expr} AS evaluation_id" in ddl
+    assert "coalesce(" in ddl
     assert "attributes_string['{test_case_key}'] AS test_case_id" in ddl
     assert "root_status LowCardinality(String)" in ddl
     assert "root_input String" in ddl
@@ -72,5 +75,8 @@ def test_trace_index_schema_is_root_span_projection():
 
 
 def test_trace_index_mv_keys_match_attribute_catalog():
-    assert spec_for_field(SpanAttributeField.EVALUATION_ID).bag_key == "nemo.experiment.id"
+    evaluation_spec = spec_for_field(SpanAttributeField.EVALUATION_NAME)
+    assert evaluation_spec.bag_key == "nemo.evaluation.name"
+    # The legacy key is still accepted on ingest so pre-rename producers keep associating.
+    assert "nemo.experiment.id" in evaluation_spec.source_keys
     assert spec_for_field(SpanAttributeField.TEST_CASE_ID).bag_key == "nemo.test_case.id"
diff --git a/services/intake/tests/test_spans_otlp.py b/services/intake/tests/test_spans_otlp.py
index fdef1edd1b..c045fe34f2 100644
--- a/services/intake/tests/test_spans_otlp.py
+++ b/services/intake/tests/test_spans_otlp.py
@@ -187,10 +187,10 @@ def test_span_to_domain_promotes_pydantic_ai_tool_arguments_and_response():
 
 
 def test_span_to_domain_carries_evaluation_id_attribute() -> None:
-    # OTLP associates a span with an evaluation via the nemo.experiment.id attribute; the ingest path
+    # OTLP associates a span with an evaluation via the nemo.evaluation.name attribute; the ingest path
     # reads it off the built span to mark that evaluation dirty for facet denormalization.
     span = _make_span()
-    _add_string_attr(span, "nemo.experiment.id", "my-eval")
+    _add_string_attr(span, "nemo.evaluation.name", "my-eval")
     domain_span = _span_to_domain(
         workspace="default",
         span=span,
@@ -198,7 +198,23 @@ def test_span_to_domain_carries_evaluation_id_attribute() -> None:
         scope_data={},
         ingested_at=datetime(2026, 1, 1, tzinfo=timezone.utc),
     )
-    assert domain_span.attributes_string.get("nemo.experiment.id") == "my-eval"
+    assert domain_span.attributes_string.get("nemo.evaluation.name") == "my-eval"
+
+
+def test_span_to_domain_normalizes_legacy_evaluation_id_key() -> None:
+    # A pre-rename producer still emitting nemo.experiment.id must keep associating: ingest normalizes
+    # the legacy source key to the canonical nemo.evaluation.name bag key.
+    span = _make_span()
+    _add_string_attr(span, "nemo.experiment.id", "legacy-eval")
+    domain_span = _span_to_domain(
+        workspace="default",
+        span=span,
+        resource_attributes={},
+        scope_data={},
+        ingested_at=datetime(2026, 1, 1, tzinfo=timezone.utc),
+    )
+    assert domain_span.attributes_string.get("nemo.evaluation.name") == "legacy-eval"
+    assert "nemo.experiment.id" not in domain_span.attributes_string
 
 
 def test_span_to_domain_without_evaluation_attribute_has_no_id() -> None:
@@ -209,7 +225,7 @@ def test_span_to_domain_without_evaluation_attribute_has_no_id() -> None:
         scope_data={},
         ingested_at=datetime(2026, 1, 1, tzinfo=timezone.utc),
     )
-    assert domain_span.attributes_string.get("nemo.experiment.id") is None
+    assert domain_span.attributes_string.get("nemo.evaluation.name") is None
 
 
 def _make_span(

From c9b2424a4929497a18114e29a3a63fe25c795be8 Mon Sep 17 00:00:00 2001
From: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
Date: Tue, 11 Aug 2026 13:47:59 -0600
Subject: [PATCH 2/5] update test

Signed-off-by: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
---
 .../intake/tests/integration/spans/test_clickhouse_bootstrap.py  | 1 +
 1 file changed, 1 insertion(+)

diff --git a/services/intake/tests/integration/spans/test_clickhouse_bootstrap.py b/services/intake/tests/integration/spans/test_clickhouse_bootstrap.py
index 3566631ac5..c6c8ed3ed6 100644
--- a/services/intake/tests/integration/spans/test_clickhouse_bootstrap.py
+++ b/services/intake/tests/integration/spans/test_clickhouse_bootstrap.py
@@ -40,6 +40,7 @@ def test_clickhouse_bootstrap_is_idempotent(clickhouse_client: ClickHouseSpanCli
         ("ch_trace_index_0003",),
         ("ch_trace_index_0004_nemo_keys",),
         ("ch_trace_index_0005_evaluation_id",),
+        ("ch_trace_index_0006_nemo_evaluation_name",),
     ]
 
 

From 2c81af6721fba241146dce0ca05f103125bd5487 Mon Sep 17 00:00:00 2001
From: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
Date: Thu, 13 Aug 2026 10:41:38 -0600
Subject: [PATCH 3/5] cleanup comments that mention legacy key

Signed-off-by: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
---
 docs/evaluator/experiments.mdx                              | 3 +--
 .../skills/nemo-experiments-upload/SKILL.md                 | 2 +-
 .../nemo-experiments-upload/references/troubleshooting.md   | 4 ++--
 .../skills/nemo-intake/references/ingest-formats.md         | 3 +--
 .../nemo_platform/skills/nemo-experiments-upload/SKILL.md   | 2 +-
 .../nemo-experiments-upload/references/troubleshooting.md   | 4 ++--
 .../skills/nemo-intake/references/ingest-formats.md         | 3 +--
 services/intake/README.md                                   | 6 ++----
 8 files changed, 11 insertions(+), 16 deletions(-)

diff --git a/docs/evaluator/experiments.mdx b/docs/evaluator/experiments.mdx
index 7a323db086..9a1f41701d 100644
--- a/docs/evaluator/experiments.mdx
+++ b/docs/evaluator/experiments.mdx
@@ -232,8 +232,7 @@ session with the Evaluation's identity:
   `evaluation_context` object to the ingest payload
   carrying `evaluation_id` (the Evaluation's **name**) and `test_case_id`.
 - For OpenTelemetry Protocol (OTLP), set the `nemo.evaluation.name` (the Evaluation's **name**) and
-  `nemo.test_case.id` root-span attributes. The legacy key `nemo.experiment.id` is still accepted for
-  backward compatibility.
+  `nemo.test_case.id` root-span attributes.
 
 The per-evaluator scores on the leaderboard come from **evaluator results** captured on those
 sessions, either automatically from ATIF verifier rewards or explicitly through the evaluator-results
diff --git a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/SKILL.md b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/SKILL.md
index f1c158157f..d189516892 100644
--- a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/SKILL.md
+++ b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/SKILL.md
@@ -192,7 +192,7 @@ If `run_count` is 0 after ingesting, the traces didn't associate — almost alwa
 
 - **Create before you log.** The Evaluation entity must exist before any ingest referencing it — otherwise `400`.
 - **`evaluation_id` is the Evaluation's `name`, not its entity id.** But **`experiment_ids` holds the Experiment's `id`.** Different identifiers; easy to swap.
-- **OTLP uses the attribute key `nemo.evaluation.name`** (and `nemo.test_case.id`) — set it to the Evaluation's **name** on your root span, matching the `evaluation_id` field the JSON `evaluation_context` carries on the other endpoints. The pre-rename key `nemo.experiment.id` is still accepted on ingest, so older exporters keep associating.
+- **OTLP uses the attribute key `nemo.evaluation.name`** (and `nemo.test_case.id`) — set it to the Evaluation's **name** on your root span, matching the `evaluation_id` field the JSON `evaluation_context` carries on the other endpoints.
 - **The parent lives at `/experiments`; `/experiment-groups` is a deprecated hidden alias.** Prefer `/experiments`. Evaluations are created and logged under `/evaluations`.
 - **`metadata` is `dict[str, str]`** — stringify non-string values or you'll get a `422`.
 - **ATIF and chat-completions are `extra="forbid"`** (unknown keys → 422); `evaluation_context` itself is lenient (`extra="ignore"`).
diff --git a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/references/troubleshooting.md b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/references/troubleshooting.md
index 7945cf5eb2..1546598608 100644
--- a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/references/troubleshooting.md
+++ b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-experiments-upload/references/troubleshooting.md
@@ -29,7 +29,7 @@ string — read it first.
 
 | Symptom | Cause | Fix |
 |---|---|---|
-| Ingest returned 2xx but `run_count` stays 0 | `evaluation_context` missing, or `evaluation_id` ≠ the Evaluation's name | Attach `evaluation_context`; use the Evaluation **name**. For OTLP, set the span attribute `nemo.evaluation.name` on the root span (legacy `nemo.experiment.id` still accepted) |
+| Ingest returned 2xx but `run_count` stays 0 | `evaluation_context` missing, or `evaluation_id` ≠ the Evaluation's name | Attach `evaluation_context`; use the Evaluation **name**. For OTLP, set the span attribute `nemo.evaluation.name` on the root span |
 | No scores on the evaluation | Rewards not under `extra.verifier_result.rewards` (ATIF), or wrong `data_type` (`/evaluator-results`) | ATIF: `extra.verifier_result.rewards = {criterion: value}`. Explicit: `NUMERIC`/`BOOLEAN` need `value`, `CATEGORICAL`/`TEXT` need `string_value` |
 | No cost on the rollup | The producer never emitted cost | Cost is pass-through — set `cost_usd` (chat-completions / ATIF step `metrics`) or `llm.cost.total` / `gen_ai.usage.cost` (OTLP) |
 | `503` on `GET .../evaluations/{name}` or `/sessions` | ClickHouse (telemetry store) not running | Start ClickHouse; rollups, sessions, and metric sorts/filters all need it |
@@ -39,5 +39,5 @@ string — read it first.
 
 - `evaluation_context.evaluation_id` → the Evaluation's **`name`**.
 - `experiment_ids` (on create evaluation) → a list with the Experiment's **`id`**.
-- OTLP evaluation attribute key → **`nemo.evaluation.name`** (legacy `nemo.experiment.id` still accepted; test case → `nemo.test_case.id`).
+- OTLP evaluation attribute key → **`nemo.evaluation.name`** (test case → `nemo.test_case.id`).
 - Parent → **`/experiments`** (`/experiment-groups` is a deprecated hidden alias); evaluations → **`/evaluations`**.
diff --git a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-intake/references/ingest-formats.md b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-intake/references/ingest-formats.md
index e4d5a248b3..4cc1d0939b 100644
--- a/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-intake/references/ingest-formats.md
+++ b/packages/nemo_platform_ext/src/nemo_platform_ext/skills/nemo-intake/references/ingest-formats.md
@@ -132,8 +132,7 @@ the root span:
 | Test case | **`nemo.test_case.id`** |
 
 > Set `nemo.evaluation.name` to the Evaluation's **name** (not its id), matching the `evaluation_id`
-> field used by the JSON `evaluation_context` on the other endpoints. The pre-rename key
-> `nemo.experiment.id` is still accepted on ingest, so existing exporters keep working.
+> field used by the JSON `evaluation_context` on the other endpoints.
 
 Cost/token/model attributes are read from standard GenAI / OpenInference keys (first match wins):
 
diff --git a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/SKILL.md b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/SKILL.md
index f1c158157f..d189516892 100644
--- a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/SKILL.md
+++ b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/SKILL.md
@@ -192,7 +192,7 @@ If `run_count` is 0 after ingesting, the traces didn't associate — almost alwa
 
 - **Create before you log.** The Evaluation entity must exist before any ingest referencing it — otherwise `400`.
 - **`evaluation_id` is the Evaluation's `name`, not its entity id.** But **`experiment_ids` holds the Experiment's `id`.** Different identifiers; easy to swap.
-- **OTLP uses the attribute key `nemo.evaluation.name`** (and `nemo.test_case.id`) — set it to the Evaluation's **name** on your root span, matching the `evaluation_id` field the JSON `evaluation_context` carries on the other endpoints. The pre-rename key `nemo.experiment.id` is still accepted on ingest, so older exporters keep associating.
+- **OTLP uses the attribute key `nemo.evaluation.name`** (and `nemo.test_case.id`) — set it to the Evaluation's **name** on your root span, matching the `evaluation_id` field the JSON `evaluation_context` carries on the other endpoints.
 - **The parent lives at `/experiments`; `/experiment-groups` is a deprecated hidden alias.** Prefer `/experiments`. Evaluations are created and logged under `/evaluations`.
 - **`metadata` is `dict[str, str]`** — stringify non-string values or you'll get a `422`.
 - **ATIF and chat-completions are `extra="forbid"`** (unknown keys → 422); `evaluation_context` itself is lenient (`extra="ignore"`).
diff --git a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/references/troubleshooting.md b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/references/troubleshooting.md
index 7945cf5eb2..1546598608 100644
--- a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/references/troubleshooting.md
+++ b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-experiments-upload/references/troubleshooting.md
@@ -29,7 +29,7 @@ string — read it first.
 
 | Symptom | Cause | Fix |
 |---|---|---|
-| Ingest returned 2xx but `run_count` stays 0 | `evaluation_context` missing, or `evaluation_id` ≠ the Evaluation's name | Attach `evaluation_context`; use the Evaluation **name**. For OTLP, set the span attribute `nemo.evaluation.name` on the root span (legacy `nemo.experiment.id` still accepted) |
+| Ingest returned 2xx but `run_count` stays 0 | `evaluation_context` missing, or `evaluation_id` ≠ the Evaluation's name | Attach `evaluation_context`; use the Evaluation **name**. For OTLP, set the span attribute `nemo.evaluation.name` on the root span |
 | No scores on the evaluation | Rewards not under `extra.verifier_result.rewards` (ATIF), or wrong `data_type` (`/evaluator-results`) | ATIF: `extra.verifier_result.rewards = {criterion: value}`. Explicit: `NUMERIC`/`BOOLEAN` need `value`, `CATEGORICAL`/`TEXT` need `string_value` |
 | No cost on the rollup | The producer never emitted cost | Cost is pass-through — set `cost_usd` (chat-completions / ATIF step `metrics`) or `llm.cost.total` / `gen_ai.usage.cost` (OTLP) |
 | `503` on `GET .../evaluations/{name}` or `/sessions` | ClickHouse (telemetry store) not running | Start ClickHouse; rollups, sessions, and metric sorts/filters all need it |
@@ -39,5 +39,5 @@ string — read it first.
 
 - `evaluation_context.evaluation_id` → the Evaluation's **`name`**.
 - `experiment_ids` (on create evaluation) → a list with the Experiment's **`id`**.
-- OTLP evaluation attribute key → **`nemo.evaluation.name`** (legacy `nemo.experiment.id` still accepted; test case → `nemo.test_case.id`).
+- OTLP evaluation attribute key → **`nemo.evaluation.name`** (test case → `nemo.test_case.id`).
 - Parent → **`/experiments`** (`/experiment-groups` is a deprecated hidden alias); evaluations → **`/evaluations`**.
diff --git a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-intake/references/ingest-formats.md b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-intake/references/ingest-formats.md
index e4d5a248b3..4cc1d0939b 100644
--- a/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-intake/references/ingest-formats.md
+++ b/sdk/python/nemo-platform/src/nemo_platform/skills/nemo-intake/references/ingest-formats.md
@@ -132,8 +132,7 @@ the root span:
 | Test case | **`nemo.test_case.id`** |
 
 > Set `nemo.evaluation.name` to the Evaluation's **name** (not its id), matching the `evaluation_id`
-> field used by the JSON `evaluation_context` on the other endpoints. The pre-rename key
-> `nemo.experiment.id` is still accepted on ingest, so existing exporters keep working.
+> field used by the JSON `evaluation_context` on the other endpoints.
 
 Cost/token/model attributes are read from standard GenAI / OpenInference keys (first match wins):
 
diff --git a/services/intake/README.md b/services/intake/README.md
index 7b130385eb..016e521e04 100644
--- a/services/intake/README.md
+++ b/services/intake/README.md
@@ -34,8 +34,7 @@ The Experiments feature captures evaluation runs as leaderboard rows. The flow:
 3. Log traces + evaluator results to an ingest endpoint. The Evaluation must exist first. Attach
    evaluation identity per endpoint: **ATIF/Harbor** and **chat-completions** carry it in the JSON body
    as `evaluation_context = {evaluation_id, test_case_id}`; **OTLP** carries it as root-span attributes
-   `nemo.evaluation.name` (the Evaluation name; legacy `nemo.experiment.id` still accepted) and
-   `nemo.test_case.id`.
+   `nemo.evaluation.name` (the Evaluation name) and `nemo.test_case.id`.
 4. Read the rollups — `GET /evaluations/{name}` and `.../{name}/sessions` — or view them in Studio
    (behind the `VITE_FF_EXPERIMENT` flag).
 
@@ -45,8 +44,7 @@ mapping, and a troubleshooting table, see the **`nemo-experiments-upload`** agen
 
 > Naming note: the entity/API is mid-rename. The API surfaces "Evaluation" and "Experiment Group",
 > but the stored entity is still `Experiment`. The OTLP evaluation attribute key is
-> `nemo.evaluation.name` (the pre-rename `nemo.experiment.id` is still accepted on ingest). This doc
-> follows the current in-code naming.
+> `nemo.evaluation.name`. This doc follows the current in-code naming.
 
 ## Local Development
 

From 967dc624c01594999e5970059c54fcbd92555fa8 Mon Sep 17 00:00:00 2001
From: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
Date: Thu, 13 Aug 2026 11:10:37 -0600
Subject: [PATCH 4/5] refactor(experiments): Rename producer evaluation
 parameters

Signed-off-by: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
---
 .../record_tau_airline_traces.py              | 28 ++++++-------
 .../experimentalist/atif.py                   |  6 +--
 .../experimentalist_backend.py                | 40 +++++++++----------
 .../nemo-experimentalist/tests/test_atif.py   | 10 ++---
 .../tests/test_experimentalist_backend.py     |  6 ++-
 plugins/nemo-insights/testbed/adapters.py     |  6 +--
 plugins/nemo-insights/testbed/otlp_build.py   | 20 +++++-----
 .../tests/testbed/test_otlp_build.py          | 24 +++++------
 8 files changed, 68 insertions(+), 72 deletions(-)

diff --git a/plugins/nemo-experimentalist/examples/tau3-nooa-agent/record_tau_airline_traces.py b/plugins/nemo-experimentalist/examples/tau3-nooa-agent/record_tau_airline_traces.py
index 0ceddda0a5..958dba5a67 100644
--- a/plugins/nemo-experimentalist/examples/tau3-nooa-agent/record_tau_airline_traces.py
+++ b/plugins/nemo-experimentalist/examples/tau3-nooa-agent/record_tau_airline_traces.py
@@ -43,7 +43,7 @@ def _positive_int(value: str) -> int:
     return parsed
 
 
-def _experiment_id() -> str:
+def _evaluation_name() -> str:
     stamp = datetime.now(UTC).strftime("%Y%m%d-%H%M%S")
     return f"tau3-airline-{stamp}-{secrets.token_hex(2)}"
 
@@ -96,7 +96,7 @@ def _resolve_harbor_output_dir(path: Path) -> tuple[Path, Path]:
 def _write_upload_summary(
     run_dir: Path,
     *,
-    experiment_id: str,
+    evaluation_name: str,
     workspace: str,
     agent_name: str,
     agent_version: str,
@@ -108,7 +108,7 @@ def _write_upload_summary(
     summary_path.write_text(
         json.dumps(
             {
-                "experiment_id": experiment_id,
+                "evaluation_name": evaluation_name,
                 "workspace": workspace,
                 "agent_name": agent_name,
                 "agent_version": agent_version,
@@ -133,9 +133,7 @@ async def _upload_trials(
     trials: list[TrialResult],
     *,
     workspace: str,
-    # TODO: rename experiment_id -> evaluation_name to match the ingested attribute
-    # (nemo.evaluation.name carries the Evaluation's name). Plumb the rename through callers.
-    experiment_id: str,
+    evaluation_name: str,
     agent_name: str,
     agent_version: str,
     model: str,
@@ -151,7 +149,7 @@ async def _upload_trials(
             raise RuntimeError(f"Trace {trace_id} was produced by more than one trial")
 
         attrs = {
-            "nemo.evaluation.name": experiment_id,
+            "nemo.evaluation.name": evaluation_name,
             "nemo.test_case.id": trial.task_id,
             "nemo.trial.id": trial.id,
             "gen_ai.agent.name": agent_name,
@@ -222,7 +220,7 @@ async def run(args: argparse.Namespace) -> Path:
         if not uploadable_trials:
             raise RuntimeError(f"No completed Harbor trials with trace artifacts found in {job_dir}")
 
-        experiment_id = args.experiment_id or run_dir.name
+        evaluation_name = args.evaluation_name or run_dir.name
         client = make_client(args.base_url)
         try:
             await client.workspaces.create(
@@ -234,7 +232,7 @@ async def run(args: argparse.Namespace) -> Path:
                 client,
                 uploadable_trials,
                 workspace=args.workspace,
-                experiment_id=experiment_id,
+                evaluation_name=evaluation_name,
                 agent_name=args.agent_name,
                 agent_version=args.agent_version,
                 model=args.model,
@@ -245,7 +243,7 @@ async def run(args: argparse.Namespace) -> Path:
 
         summary_path = _write_upload_summary(
             run_dir,
-            experiment_id=experiment_id,
+            evaluation_name=evaluation_name,
             workspace=args.workspace,
             agent_name=args.agent_name,
             agent_version=args.agent_version,
@@ -257,8 +255,8 @@ async def run(args: argparse.Namespace) -> Path:
         print(summary_path)
         return summary_path
 
-    experiment_id = args.experiment_id or _experiment_id()
-    run_dir = args.output.expanduser().resolve() / experiment_id
+    evaluation_name = args.evaluation_name or _evaluation_name()
+    run_dir = args.output.expanduser().resolve() / evaluation_name
     if run_dir.exists():
         raise FileExistsError(f"Output directory already exists: {run_dir}")
 
@@ -296,7 +294,7 @@ async def run(args: argparse.Namespace) -> Path:
             client,
             trials,
             workspace=args.workspace,
-            experiment_id=experiment_id,
+            evaluation_name=evaluation_name,
             agent_name=args.agent_name,
             agent_version=args.agent_version,
             model=args.model,
@@ -307,7 +305,7 @@ async def run(args: argparse.Namespace) -> Path:
 
     summary_path = _write_upload_summary(
         run_dir,
-        experiment_id=experiment_id,
+        evaluation_name=evaluation_name,
         workspace=args.workspace,
         agent_name=args.agent_name,
         agent_version=args.agent_version,
@@ -340,7 +338,7 @@ def parse_args() -> argparse.Namespace:
     parser.add_argument("--user-model", default=os.environ.get("TAU2_USER_MODEL", DEFAULT_MODEL))
     parser.add_argument("--agent-name", default=DEFAULT_AGENT_NAME)
     parser.add_argument("--agent-version", default=DEFAULT_AGENT_VERSION)
-    parser.add_argument("--experiment-id")
+    parser.add_argument("--evaluation-name")
     parser.add_argument("--expected-task-count", type=_positive_int, default=20)
     parser.add_argument(
         "--task-id",
diff --git a/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/atif.py b/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/atif.py
index cfa20e450d..dd2c4e75ce 100644
--- a/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/atif.py
+++ b/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/atif.py
@@ -51,7 +51,7 @@ def read_session_id(ref: ResourceRef) -> str:
 def build_ingest_payload(
     ref: ResourceRef,
     *,
-    experiment_id: str,
+    evaluation_name: str,
     task_id: str,
     agent_attrs: dict[str, str],
 ) -> dict[str, Any]:
@@ -66,7 +66,7 @@ def build_ingest_payload(
 
     Args:
         ref(ResourceRef): Resource reference to the ATIF trajectory file.
-        experiment_id(str): Evaluation name, used as ``evaluation_context.evaluation_id``.
+        evaluation_name(str): Evaluation name, used as ``evaluation_context.evaluation_id``.
         task_id(str): Test case id, used as ``evaluation_context.test_case_id``.
         agent_attrs(dict[str, str]): OTLP-style agent attributes used as fallbacks.
 
@@ -75,7 +75,7 @@ def build_ingest_payload(
     """
     trajectory = _load(ref)
     trajectory["evaluation_context"] = {
-        "evaluation_id": experiment_id,
+        "evaluation_id": evaluation_name,
         "test_case_id": task_id,
     }
     agent = trajectory.get("agent")
diff --git a/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experimentalist_backend.py b/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experimentalist_backend.py
index 957d247eea..a9cfebf172 100644
--- a/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experimentalist_backend.py
+++ b/plugins/nemo-experimentalist/src/nemo_experimentalist_plugin/experimentalist/experimentalist_backend.py
@@ -58,9 +58,7 @@ async def _upload_trace_otlp(
     workspace: str,
     ref: ResourceRef,
     *,
-    # TODO: rename experiment_id -> evaluation_name to match the ingested attribute
-    # (nemo.evaluation.name carries the Evaluation's name). Plumb the rename through callers.
-    experiment_id: str,
+    evaluation_name: str,
     trial_id: str,
     task_id: str,
     extra_attrs: dict[str, str] | None = None,
@@ -69,7 +67,7 @@ async def _upload_trace_otlp(
 
     path = Path(urlparse(ref.uri).path)
     attrs: dict[str, str] = {
-        "nemo.evaluation.name": experiment_id,
+        "nemo.evaluation.name": evaluation_name,
         "nemo.test_case.id": task_id,
         "nemo.trial.id": trial_id,
         **(extra_attrs or {}),
@@ -89,7 +87,7 @@ async def _upload_trace_atif(
     workspace: str,
     ref: ResourceRef,
     *,
-    experiment_id: str,
+    evaluation_name: str,
     task_id: str,
     extra_attrs: dict[str, str] | None = None,
 ) -> None:
@@ -102,7 +100,7 @@ async def _upload_trace_atif(
 
     payload = build_ingest_payload(
         ref,
-        experiment_id=experiment_id,
+        evaluation_name=evaluation_name,
         task_id=task_id,
         agent_attrs=extra_attrs or {},
     )
@@ -208,16 +206,16 @@ async def persist_evaluation(
     ) -> None:
         """Persist traces and metrics for a completed evaluation.
 
-        Derives experiment_id internally from candidate × split, ensuring the
-        Experiment entity exists before stamping traces.
+        Derives the Evaluation name internally from candidate × split, ensuring
+        the Evaluation exists before stamping traces.
         """
         ...
 
     @abstractmethod
-    async def get_experiment_id(self, *, workspace: str, candidate: Candidate, split: str) -> str:
-        """Best-effort native Experiment id for *candidate* × *split* in *workspace*.
+    async def get_evaluation_name(self, *, workspace: str, candidate: Candidate, split: str) -> str:
+        """Best-effort Evaluation name for *candidate* × *split* in *workspace*.
 
-        Returns "" when there is no projection (offline) or it fails — the id only
+        Returns "" when there is no projection (offline) or it fails — the name only
         tags Intake resource attributes, so a run must not break on it.
         """
         ...
@@ -632,9 +630,9 @@ async def persist_evaluation(
         if self.client is None:
             return  # pure-offline run: traces stay on local disk
 
-        # Derive experiment_id internally (ensures entity exists, returns deterministic name)
-        experiment_id = await self.get_experiment_id(workspace=workspace, candidate=candidate, split=split)
-        if not experiment_id:
+        # Derive the Evaluation name internally (ensures the entity exists and returns a deterministic name).
+        evaluation_name = await self.get_evaluation_name(workspace=workspace, candidate=candidate, split=split)
+        if not evaluation_name:
             return  # projection failed (shouldn't happen, but defensive)
 
         agent_attrs = {
@@ -651,13 +649,13 @@ async def persist_evaluation(
                 continue
             try:
                 await self._persist_trial(
-                    trial, workspace=workspace, experiment_id=experiment_id, agent_attrs=agent_attrs
+                    trial, workspace=workspace, evaluation_name=evaluation_name, agent_attrs=agent_attrs
                 )
             except Exception as exc:  # noqa: BLE001 - Intake persistence is best-effort
                 logger.warning(f"[INTAKE] persist_evaluation failed for trial {trial.id}: {exc}")
 
     async def _persist_trial(
-        self, trial: TrialResult, *, workspace: str, experiment_id: str, agent_attrs: dict[str, str]
+        self, trial: TrialResult, *, workspace: str, evaluation_name: str, agent_attrs: dict[str, str]
     ) -> None:
         assert trial.trace is not None
         assert self.client is not None
@@ -666,7 +664,7 @@ async def _persist_trial(
             trace_id = uri.removeprefix("intake://traces/")
             trace = await self._retrieve_trace_with_retry(trace_id, workspace=workspace)
             ctx = getattr(trace, "evaluation_context", None)
-            if ctx is None or getattr(ctx, "evaluation_id", None) != experiment_id:
+            if ctx is None or getattr(ctx, "evaluation_id", None) != evaluation_name:
                 rows: list[dict] = []
                 async for span in self.client.intake.spans.list(
                     workspace=workspace,
@@ -676,7 +674,7 @@ async def _persist_trial(
                 ):
                     rows.append(span.model_dump(mode="json", exclude_none=True))
                 attrs = {
-                    "nemo.evaluation.name": experiment_id,
+                    "nemo.evaluation.name": evaluation_name,
                     "nemo.test_case.id": trial.task_id,
                     "nemo.trial.id": trial.id,
                     **agent_attrs,
@@ -704,7 +702,7 @@ async def _persist_trial(
                     self.client,
                     workspace,
                     trial.trace,
-                    experiment_id=experiment_id,
+                    evaluation_name=evaluation_name,
                     task_id=trial.task_id,
                     extra_attrs=agent_attrs,
                 )
@@ -713,7 +711,7 @@ async def _persist_trial(
                     self.client,
                     workspace,
                     trial.trace,
-                    experiment_id=experiment_id,
+                    evaluation_name=evaluation_name,
                     trial_id=trial.id,
                     task_id=trial.task_id,
                     extra_attrs=agent_attrs,
@@ -757,7 +755,7 @@ async def _retrieve_trace_with_retry(
                     delay *= 2  # Exponential backoff
         raise last_exc
 
-    async def get_experiment_id(self, *, workspace: str, candidate: Candidate, split: str) -> str:
+    async def get_evaluation_name(self, *, workspace: str, candidate: Candidate, split: str) -> str:
         if self.client is None:
             return ""
         mirror = self._mirrors.get(workspace)
diff --git a/plugins/nemo-experimentalist/tests/test_atif.py b/plugins/nemo-experimentalist/tests/test_atif.py
index 6c755ba14f..0c64f073ae 100644
--- a/plugins/nemo-experimentalist/tests/test_atif.py
+++ b/plugins/nemo-experimentalist/tests/test_atif.py
@@ -70,7 +70,7 @@ def test_read_session_id_raises_on_empty_string(tmp_path):
 def test_build_ingest_payload_stamps_evaluation_context(tmp_path):
     payload = build_ingest_payload(
         _ref(tmp_path, _trajectory()),
-        experiment_id="exp-1",
+        evaluation_name="exp-1",
         task_id="tau3-airline/case-a",
         agent_attrs={},
     )
@@ -83,7 +83,7 @@ def test_build_ingest_payload_stamps_evaluation_context(tmp_path):
 def test_build_ingest_payload_preserves_producer_fields(tmp_path):
     payload = build_ingest_payload(
         _ref(tmp_path, _trajectory()),
-        experiment_id="exp-1",
+        evaluation_name="exp-1",
         task_id="case-a",
         agent_attrs={},
     )
@@ -96,7 +96,7 @@ def test_build_ingest_payload_fills_only_blank_agent_fields(tmp_path):
     # Producer set name and version; model_name is absent.
     payload = build_ingest_payload(
         _ref(tmp_path, _trajectory()),
-        experiment_id="exp-1",
+        evaluation_name="exp-1",
         task_id="case-a",
         agent_attrs=AGENT_ATTRS,
     )
@@ -110,7 +110,7 @@ def test_build_ingest_payload_fills_missing_agent_block(tmp_path):
     del trajectory["agent"]
     payload = build_ingest_payload(
         _ref(tmp_path, trajectory),
-        experiment_id="exp-1",
+        evaluation_name="exp-1",
         task_id="case-a",
         agent_attrs=AGENT_ATTRS,
     )
@@ -124,7 +124,7 @@ def test_build_ingest_payload_fills_missing_agent_block(tmp_path):
 def test_build_ingest_payload_does_not_inject_verifier_result(tmp_path):
     payload = build_ingest_payload(
         _ref(tmp_path, _trajectory()),
-        experiment_id="exp-1",
+        evaluation_name="exp-1",
         task_id="case-a",
         agent_attrs=AGENT_ATTRS,
     )
diff --git a/plugins/nemo-experimentalist/tests/test_experimentalist_backend.py b/plugins/nemo-experimentalist/tests/test_experimentalist_backend.py
index 241092bdb8..20cd72a72f 100644
--- a/plugins/nemo-experimentalist/tests/test_experimentalist_backend.py
+++ b/plugins/nemo-experimentalist/tests/test_experimentalist_backend.py
@@ -397,7 +397,7 @@ async def test_upload_trace_atif_posts_to_the_atif_ingest_endpoint(tmp_path):
         cast(Any, client),
         "ws-1",
         _atif_ref(tmp_path),
-        experiment_id="exp-1",
+        evaluation_name="exp-1",
         task_id="case-a",
         extra_attrs={"gen_ai.request.model": "gpt-5-mini"},
     )
@@ -416,6 +416,8 @@ async def test_upload_trace_atif_sends_json_not_protobuf(tmp_path):
     from nemo_experimentalist_plugin.experimentalist.experimentalist_backend import _upload_trace_atif
 
     client = _RecordingClient()
-    await _upload_trace_atif(cast(Any, client), "ws-1", _atif_ref(tmp_path), experiment_id="exp-1", task_id="case-a")
+    await _upload_trace_atif(
+        cast(Any, client), "ws-1", _atif_ref(tmp_path), evaluation_name="exp-1", task_id="case-a"
+    )
     assert client.posts[0]["content"] is None
     assert client.posts[0]["body"] is not None
diff --git a/plugins/nemo-insights/testbed/adapters.py b/plugins/nemo-insights/testbed/adapters.py
index 77a32e47f2..d913287040 100644
--- a/plugins/nemo-insights/testbed/adapters.py
+++ b/plugins/nemo-insights/testbed/adapters.py
@@ -200,7 +200,7 @@ async def produce(self) -> dict[str, object]:
         try:
             for sim in sims:
                 task = tasks.get(str(sim.get("task_id")))
-                session_id = session_id_for(sim, experiment_id=run_id)
+                session_id = session_id_for(sim, evaluation_name=run_id)
                 trace_id = trace_id_for(session_id)
                 # Stamp spans at ingest time (Intake drops spans dated outside its
                 # retention window); one base shared by a sim's realistic + oracle twins.
@@ -211,7 +211,7 @@ async def produce(self) -> dict[str, object]:
                     agent_name=agent,
                     agent_version=version,
                     session_id=session_id,
-                    experiment_id=run_id,
+                    evaluation_name=run_id,
                     task=task,
                     include_rewards=False,
                     agent_llm=agent_llm,
@@ -224,7 +224,7 @@ async def produce(self) -> dict[str, object]:
                         agent_name=agent,
                         agent_version=version,
                         session_id=session_id,
-                        experiment_id=run_id,
+                        evaluation_name=run_id,
                         task=task,
                         include_rewards=True,
                         agent_llm=agent_llm,
diff --git a/plugins/nemo-insights/testbed/otlp_build.py b/plugins/nemo-insights/testbed/otlp_build.py
index eb838c476a..657b97fc6a 100644
--- a/plugins/nemo-insights/testbed/otlp_build.py
+++ b/plugins/nemo-insights/testbed/otlp_build.py
@@ -41,18 +41,18 @@ def _slug(value: object) -> str:
     return _SLUG.sub("-", str(value))
 
 
-def session_id_for(sim: dict[str, Any], *, experiment_id: str) -> str:
-    """The per-run, per-sim session id: ``<experiment_id>-task<task_id>-t<trial>``.
+def session_id_for(sim: dict[str, Any], *, evaluation_name: str) -> str:
+    """The per-run, per-sim session id: ``<evaluation_name>-task<task_id>-t<trial>``.
 
-    Scoped by ``experiment_id`` (the run id) so every id derived from it — the span
+    Scoped by ``evaluation_name`` (the run id) so every id derived from it — the span
     id, trace id, and the evaluator-results id — is unique per run. Runs that
     re-cover the same task in a shared workspace no longer collide (which would
-    cross-link spans and overwrite one run's scores with another's). ``experiment_id``
+    cross-link spans and overwrite one run's scores with another's). ``evaluation_name``
     already carries the agent/base name as its prefix, so the agent stays visible,
     and the realistic + oracle twins of one sim in one run still share this id.
     """
     trial = sim.get("trial") or 0
-    return f"{_slug(experiment_id)}-task{_slug(sim.get('task_id', ''))}-t{trial}"
+    return f"{_slug(evaluation_name)}-task{_slug(sim.get('task_id', ''))}-t{trial}"
 
 
 def _span_id(session_id: str, index: int) -> str:
@@ -78,9 +78,7 @@ def sim_to_spans(
     agent_name: str,
     agent_version: str,
     session_id: str,
-    # TODO: rename experiment_id -> evaluation_name to match the ingested attribute
-    # (nemo.evaluation.name carries the Evaluation's name). Plumb the rename through callers.
-    experiment_id: str,
+    evaluation_name: str,
     task: dict[str, Any] | None = None,
     include_rewards: bool,
     agent_llm: str | None = None,
@@ -97,9 +95,9 @@ def sim_to_spans(
     task's ``evaluation_criteria`` only appear when it is set (the realistic twin
     withholds them so the eval is unaided).
 
-    ``experiment_id`` is stamped on every span as ``nemo.evaluation.name`` (with
+    ``evaluation_name`` is stamped on every span as ``nemo.evaluation.name`` (with
     the sim's task id as ``nemo.test_case.id``) so a run's spans are queryable
-    back via the spans filter ``{"evaluation_id": experiment_id}`` — the per-run
+    back via the spans filter ``{"evaluation_id": evaluation_name}`` — the per-run
     scope that lets many runs share one workspace.
 
     ``base_ns`` seeds the per-span monotonic clock; it must be near ingest time
@@ -115,7 +113,7 @@ def _common() -> dict[str, Any]:
         return {
             "gen_ai.conversation.id": session_id,
             "session.id": session_id,
-            "nemo.evaluation.name": experiment_id,
+            "nemo.evaluation.name": evaluation_name,
             "nemo.test_case.id": test_case_id,
         }
 
diff --git a/plugins/nemo-insights/tests/testbed/test_otlp_build.py b/plugins/nemo-insights/tests/testbed/test_otlp_build.py
index 19767cc98e..43aebbc552 100644
--- a/plugins/nemo-insights/tests/testbed/test_otlp_build.py
+++ b/plugins/nemo-insights/tests/testbed/test_otlp_build.py
@@ -16,7 +16,7 @@ def _spans(sim, *, include_rewards=False, task=None, agent_llm="openai/m"):
         agent_name="tau2-airline",
         agent_version="v1",
         session_id=_SESSION,
-        experiment_id="tau2-airline-20260626-000000-abcd",
+        evaluation_name="tau2-airline-20260626-000000-abcd",
         task=task,
         include_rewards=include_rewards,
         agent_llm=agent_llm,
@@ -92,7 +92,7 @@ def test_base_ns_is_honored_when_provided():
         agent_name="a",
         agent_version="v",
         session_id="s",
-        experiment_id="run-x",
+        evaluation_name="run-x",
         task=None,
         include_rewards=False,
         agent_llm="m",
@@ -107,24 +107,24 @@ def test_base_ns_is_honored_when_provided():
 
 def test_session_id_basic():
     sim = {"task_id": "7", "trial": 0}
-    assert session_id_for(sim, experiment_id="tau2-airline-x") == "tau2-airline-x-task7-t0"
+    assert session_id_for(sim, evaluation_name="tau2-airline-x") == "tau2-airline-x-task7-t0"
 
 
-def test_session_id_slugs_experiment_and_task():
+def test_session_id_slugs_evaluation_and_task():
     sim = {"task_id": "set[1]/odd", "trial": 0}
-    assert session_id_for(sim, experiment_id="tau2/airline") == "tau2-airline-taskset-1-odd-t0"
+    assert session_id_for(sim, evaluation_name="tau2/airline") == "tau2-airline-taskset-1-odd-t0"
 
 
 def test_session_id_trial_none_defaults_to_zero():
-    assert session_id_for({"task_id": "0", "trial": None}, experiment_id="a").endswith("-t0")
+    assert session_id_for({"task_id": "0", "trial": None}, evaluation_name="a").endswith("-t0")
 
 
 def test_session_id_is_run_unique():
     sim = {"task_id": "7", "trial": 0}
-    a = session_id_for(sim, experiment_id="run-A")
-    b = session_id_for(sim, experiment_id="run-B")
+    a = session_id_for(sim, evaluation_name="run-A")
+    b = session_id_for(sim, evaluation_name="run-B")
     assert a != b  # different runs -> different session ids
-    assert a == session_id_for(sim, experiment_id="run-A")  # same run -> stable (twins correlate)
+    assert a == session_id_for(sim, evaluation_name="run-A")  # same run -> stable (twins correlate)
 
 
 # --- AGENT root ------------------------------------------------------------
@@ -345,14 +345,14 @@ def test_task_evaluation_criteria_gated_by_include_rewards():
     assert oracle_task["evaluation_criteria"]["actions"][0]["name"] == "book_reservation"
 
 
-def test_every_span_carries_experiment_and_test_case_tags():
+def test_every_span_carries_evaluation_and_test_case_tags():
     sim = _basic_sim([{"role": "user", "content": "hi"}], task_id="7")
     spans = sim_to_spans(
         sim,
         agent_name="tau2-airline",
         agent_version="v1",
         session_id=_SESSION,
-        experiment_id="tau2-airline-20260626-000000-abcd",
+        evaluation_name="tau2-airline-20260626-000000-abcd",
         include_rewards=True,
     )
     assert spans  # at least the AGENT root + EVALUATOR
@@ -367,7 +367,7 @@ def test_tags_identical_across_realistic_and_oracle_twins():
         agent_name="tau2-airline",
         agent_version="v1",
         session_id=_SESSION,
-        experiment_id="run-1",
+        evaluation_name="run-1",
     )
     realistic = sim_to_spans(sim, include_rewards=False, **common)
     oracle = sim_to_spans(sim, include_rewards=True, **common)

From dda01b6ab212d473fc6b5808641fa27622fbcd17 Mon Sep 17 00:00:00 2001
From: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
Date: Thu, 13 Aug 2026 14:37:07 -0600
Subject: [PATCH 5/5] lint

Signed-off-by: shanaiabuggy <59746633+shanaiabuggy@users.noreply.github.com>
---
 .../tests/test_experimentalist_backend.py                     | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/plugins/nemo-experimentalist/tests/test_experimentalist_backend.py b/plugins/nemo-experimentalist/tests/test_experimentalist_backend.py
index 20cd72a72f..22275bb0b8 100644
--- a/plugins/nemo-experimentalist/tests/test_experimentalist_backend.py
+++ b/plugins/nemo-experimentalist/tests/test_experimentalist_backend.py
@@ -416,8 +416,6 @@ async def test_upload_trace_atif_sends_json_not_protobuf(tmp_path):
     from nemo_experimentalist_plugin.experimentalist.experimentalist_backend import _upload_trace_atif
 
     client = _RecordingClient()
-    await _upload_trace_atif(
-        cast(Any, client), "ws-1", _atif_ref(tmp_path), evaluation_name="exp-1", task_id="case-a"
-    )
+    await _upload_trace_atif(cast(Any, client), "ws-1", _atif_ref(tmp_path), evaluation_name="exp-1", task_id="case-a")
     assert client.posts[0]["content"] is None
     assert client.posts[0]["body"] is not None