You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 0affe52
Browse filesBrowse the repository at this point in the historyBrowse files
fix(model-capabilities): humanize databricks goose model names (#7135)
🤖
## Summary
- add curated human-readable labels for Databricks Goose models that
otherwise render as fully qualified identifiers
- render `data_workflow_tools.goose.goose-glm-5-3` as `GLM-5.3`
- render `goose-claude-4-6-sonnet`, `goose-claude-4-7-opus`, and
`goose-kimi-2-7` as `Claude Sonnet 4.6`, `Claude Opus 4.7`, and `Kimi
2.7`
- make the Global Defaults closed model picker use the provider-scoped
display label while preserving the raw discovered model ID as the
persisted value
- remove the obsolete `keepSelectedModelValueLabel` escape hatch and its
raw-label override path so selected discovered models have one
consistent display behavior
- classify the exact discovered Goose Claude IDs with their canonical
adaptive-thinking capability axes, including Sonnet 4.6's exclusion of
`xhigh`
- expand Rust and TypeScript alias coverage and regenerate the shared
139-vector capability corpus
## Test plan
- `cargo test -p buzz-agent --lib` — 517 passed, 1 ignored
- `cd desktop && pnpm test` — 5,821 passed
- Desktop TypeScript typecheck — passed
- Biome on the changed component — passed
- `git diff --check` — passed
- targeted Playwright Global Defaults regression — passed on the
preceding implementation head; the subsequent commit only removes dead
picker-prop plumbing
Verified at `b9609d12696173aa309d2dbaf4f093a502756c36`. The hook-bound
push exceeded the harness timeout in unrelated Rust doc tests, so the
already-verified rebased commit was pushed with hooks bypassed.
Follow-up to #6955.
---------
Signed-off-by: Kalvin Chau <[email protected]>
Co-authored-by: am <6e30cd56c30e030cd31bb0939b94a7c257c9a09d5ba2d92cf2735da45629f248@buzz.block.builderlab.xyz>
Q::Vector{id:"dbv2-databricks-prefix-probe",provider:"databricks_v2",raw_model_id:"databricks-claude-opus-4-7",note:Some("Probes stripping of the databricks- catalog prefix.")},
624
624
Q::Vector{id:"dbv2-goose-claude-prefix-probe",provider:"databricks_v2",raw_model_id:"goose-claude-fable-5",note:Some("Probes stripping of the goose- catalog prefix.")},
625
+
Q::Vector{id:"dbv2-goose-claude-4-6-sonnet-alias-probe",provider:"databricks_v2",raw_model_id:"goose-claude-4-6-sonnet",note:Some("Probes the discovered Goose Sonnet 4.6 endpoint spelling and label.")},
626
+
Q::Vector{id:"dbv2-goose-claude-4-7-opus-alias-probe",provider:"databricks_v2",raw_model_id:"goose-claude-4-7-opus",note:Some("Probes the discovered Goose Opus 4.7 endpoint spelling and label.")},
625
627
Q::Vector{id:"dbv2-team-prefix-probe",provider:"databricks_v2",raw_model_id:"team-x-claude-opus-4-7",note:Some("Probes stripping of a team-x- catalog prefix.")},
626
628
Q::Vector{id:"dbv2-consolidated-llama-substring-probe",provider:"databricks_v2",raw_model_id:"consolidated-llama",note:Some("Probes a name where a code word ('sol') appears only as a substring, not a boundary-aligned segment.")},
627
629
Q::Vector{id:"dbv2-terraform-coder-substring-probe",provider:"databricks_v2",raw_model_id:"terraform-coder",note:Some("Probes a name where a code word ('terra') is only a segment prefix, not a full segment.")},
@@ -638,6 +640,7 @@ mod tests {
638
640
Q::Vector{id:"dbv2-goose-claude-opus-5-alias-probe",provider:"databricks_v2",raw_model_id:"goose-claude-opus-5",note:Some("Probes a prefixed alias of the Databricks Opus 5 endpoint.")},
639
641
Q::Vector{id:"dbv2-claude-sonnet-5-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-claude-sonnet-5",note:Some("Probes the canonical Databricks Sonnet 5 endpoint record.")},
640
642
Q::Vector{id:"dbv2-goose-claude-sonnet-5-alias-probe",provider:"databricks_v2",raw_model_id:"goose-claude-sonnet-5",note:Some("Probes a prefixed alias of the Databricks Sonnet 5 endpoint.")},
643
+
Q::Vector{id:"dbv2-kimi-2-7-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-kimi-2-7",note:Some("Probes the canonical Databricks Kimi 2.7 endpoint record.")},
641
644
Q::Vector{id:"dbv2-kimi-k3-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-kimi-k3",note:Some("Probes the canonical Databricks Kimi K3 endpoint record.")},
642
645
Q::Vector{id:"dbv2-goose-kimi-k3-alias-probe",provider:"databricks_v2",raw_model_id:"goose-kimi-k3",note:Some("Probes a prefixed alias of the Databricks Kimi K3 endpoint.")},
643
646
Q::Vector{id:"resolver-prefixed-alias-probe",provider:"databricks_v2",raw_model_id:"team-x-databricks-gpt-5-4-mini",note:Some("Probes a prefixed alias of an exact-record id (raw exact key differs).")},
@@ -728,6 +731,7 @@ mod tests {
728
731
Q::Vector{id:"dbv2-gemini-3-pro-image-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-gemini-3-pro-image",note:Some("Probes the Gemini 3 Pro Image endpoint record and label.")},
729
732
Q::Vector{id:"dbv2-deepseek-v4-flash-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-deepseek-v4-flash-0731",note:Some("Probes the DeepSeek V4 Flash endpoint record and label.")},
730
733
Q::Vector{id:"dbv2-deepseek-v4-pro-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-deepseek-v4-pro-0813",note:Some("Probes the DeepSeek V4 Pro endpoint record and label.")},
734
+
Q::Vector{id:"dbv2-glm-5-3-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-glm-5-3",note:Some("Probes the GLM-5.3 endpoint record and label.")},
731
735
Q::Vector{id:"dbv2-glm-5-3-flash-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-glm-5-3-flash",note:Some("Probes the GLM-5.3 Flash endpoint record and label.")},
732
736
Q::Vector{id:"dbv2-grok-4-6-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-grok-4-6",note:Some("Probes the Grok 4.6 endpoint record and label.")},
733
737
Q::Vector{id:"dbv2-llama-4-maverick-exact-record-probe",provider:"databricks_v2",raw_model_id:"databricks-llama-4-maverick",note:Some("Probes the Llama 4 Maverick endpoint record and label.")},
@@ -839,7 +843,7 @@ mod tests {
839
843
}
840
844
841
845
#[test]
842
-
fncorpus_has_exactly_135_executable_vectors(){
846
+
fncorpus_has_exactly_139_executable_vectors(){
843
847
// Locks the vector count so a silent INPUTS edit can't quietly drop
844
848
// coverage; must equal the gate in the TS harness
845
849
// (modelCapabilitiesCorpus.test.mjs).
@@ -848,7 +852,7 @@ mod tests {
848
852
.filter(|q| matches!(q,Q::Vector{ .. }))
849
853
.count();
850
854
assert_eq!(
851
-
vectors,135,
855
+
vectors,139,
852
856
"corpus executable-vector count changed; update this gate deliberately"
Copy file name to clipboardExpand all lines: desktop/src/features/agents/AGENTS.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -281,7 +281,7 @@ with a TypeScript lookup table or an id comparison in a component.
281
281
refresh only local persona/team/managed-agent caches; they must never
282
282
invalidate the remote relay directory.
283
283
284
-
17.**Databricks model discovery has one shared catalog authority.** Desktop and ACP call the shared `buzz-agent` discovery library; Desktop passes the effective merged `DATABRICKS_MODEL_FILTER` explicitly, and the library applies it to raw workspace endpoint IDs and Unity Catalog model-service FQNs after the additive union. A successful filtered-empty catalog is authoritative: it stays empty, disables switching, and never falls through to configured or known-model fallback. UC FQNs are catalog data and always use the MLflow Chat Completions route, regardless of family-looking text in their components.
284
+
17.**Databricks model discovery has one shared catalog authority.** Desktop and ACP call the shared `buzz-agent` discovery library; Desktop passes the effective merged `DATABRICKS_MODEL_FILTER` explicitly, and the library applies it to raw workspace endpoint IDs and Unity Catalog model-service FQNs after the additive union. A successful filtered-empty catalog is authoritative: it stays empty, disables switching, and never falls through to configured or known-model fallback. UC FQNs are catalog data and always use the MLflow Chat Completions route, regardless of family-looking text in their components. Global Defaults preserves the discovered model ID as the selected value while its closed trigger renders the provider-scoped display label; do not force the raw persisted ID over that label.
Copy file name to clipboardExpand all lines: scripts/model-capabilities.json
+73Lines changed: 73 additions & 0 deletions
Original file line number
Diff line number
Diff line change
@@ -601,6 +601,24 @@
601
601
"_reconciliation_note": "models.dev advertises reasoning_options=[{\"type\":\"budget_tokens\",\"min\":1024}]. This is a different capability axis (extended thinking token budget), not an effort-level selector. No effort divergence to reconcile — efforts for this model come from the anthropic family rule (anthropic-adaptive-xhigh-opus-4-7).",
"_provenance": "all axes copied from equivalent registry endpoint databricks-claude-opus-4-7 for the discovered Goose endpoint id",
620
+
"_source": "registry_labels"
621
+
},
604
622
{
605
623
"provider": "databricks_v2",
606
624
"raw_model_id": "databricks-gpt-5-6-luna",
@@ -761,6 +779,23 @@
761
779
"_provenance": "all axes materialized from family:anthropic-adaptive-no-xhigh-sonnet-4-6",
762
780
"_source": "registry_labels"
763
781
},
782
+
{
783
+
"provider": "databricks_v2",
784
+
"raw_model_id": "goose-claude-4-6-sonnet",
785
+
"registry_label": "Claude Sonnet 4.6",
786
+
"thinking_mode": "adaptive",
787
+
"supported_efforts": [
788
+
"low",
789
+
"medium",
790
+
"high",
791
+
"max"
792
+
],
793
+
"default_effort": "high",
794
+
"databricks_v2_wire_route": "anthropic-messages",
795
+
"normalization_policy": "none",
796
+
"_provenance": "all axes copied from equivalent registry endpoint databricks-claude-sonnet-4-6 for the discovered Goose endpoint id",
797
+
"_source": "registry_labels"
798
+
},
764
799
{
765
800
"provider": "databricks_v2",
766
801
"raw_model_id": "databricks-gemini-2-5-flash",
@@ -1031,6 +1066,25 @@
1031
1066
"_reconciliation_note": "The exact databricks endpoint is absent from the models.dev databricks catalog, but the first-party deepseek entry advertises a reasoning toggle plus effort [high, max]. Per the Kimi-K3 precedent, adopt the model-level effort list despite endpoint absence. The toggle is not representable on the MLflow Chat request schema (reasoning_effort only), so thinking_mode stays none and default_effort is null — mirroring the kimi-k3 record.",
"_provenance": "all axes materialized from provider fallback (databricks_v2/concrete_unknown)",
1086
+
"_source": "registry_labels"
1087
+
},
1034
1088
{
1035
1089
"provider": "databricks_v2",
1036
1090
"raw_model_id": "databricks-glm-5-3-flash",
@@ -1385,6 +1439,25 @@
1385
1439
"_reconciliation_note": "models.dev advertises toggleable reasoning with [low, high, max], but no default. The Databricks endpoint is absent from its provider catalog, so retain the established MLflow Chat route, adopt the upstream capability set, and leave the effort unset rather than invent a Databricks default. The MLflow Chat request schema cannot express a reasoning toggle (only reasoning_effort is representable), so thinking_mode maps to none rather than the upstream toggle, matching the kimi-k2-7-code precedent.",
0 commit comments