✅ Deploy Preview for monarch-app ready!

Name	Link
🔨 Latest commit	2a26c41
🔍 Latest deploy log	https://app.netlify.com/projects/monarch-app/deploys/6a4f0810c159f90008765168
😎 Deploy Preview	https://deploy-preview-1352--monarch-app.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

To edit notification comments on pull requests, go to your Netlify project configuration.

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 83.17%. Comparing base (f28d52c) to head (2a26c41).

@@           Coverage Diff           @@
##             main    #1352   +/-   ##
=======================================
  Coverage   83.17%   83.17%           
=======================================
  Files         126      126           
  Lines        6617     6617           
=======================================
  Hits         5504     5504           
  Misses       1113     1113           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

Review: MEDIC/CTD drug-indication section + enriched details modal (#1348)

Nice, focused PR — un-hiding the ChemicalOrDrugOrTreatmentToDiseaseOrPhenotypicFeatureAssociation section and filling out SectionAssociationDetails.vue is a clean way to surface data that was already flowing through the pipeline but hidden from users.

Overview

• association_type_mappings.yaml: renames "Medical Action" → "Indications"/"Treatments" and fills in subject_category/object_category (previously omitted).
• use-association-categories.ts: empties HIDDEN_CATEGORIES, un-hiding the section.
• SectionAssociationDetails.vue: adds Knowledge Level, Provenance (via new retrievalSources.ts helper), Aggregator Knowledge Source, Original Predicate, Qualifiers, Frequency Data, FDA Adverse Event Level, and Negated detail blocks.
• New frontend/src/util/retrievalSources.ts + unit tests for parsing the nested sources provenance chain.
• Backend fixtures updated for the label rename.

Code quality / style

• The Vue additions follow the file's existing conventions well: same AppDetail/AppFlex patterns, same v-if-to-hide idiom already used for supporting_text, icons already registered in global/icons.ts, and the new qualifierRows/hasFrequencyData computeds read cleanly.
• retrievalSources.ts is a good example of defensive parsing (try/catch per entry, skips malformed data) with tests covering the nested-chain, empty, and malformed cases.
• Minor nit: const HIDDEN_CATEGORIES = new Set<string>([]); — since it's always empty now, new Set<string>() (no array arg) reads a bit more clearly, though this is trivial.

Potential issues

• object_category: "biolink:Disease" on the new mapping entry doesn't cover PhenotypicFeature. I traced how subject_category/object_category are consumed (AssociationTypeMappings.get_traversable_associations in association_type_utils.py, used by the entity-grid/traversable-associations/{entity_category} endpoint) — it does an exact-match against entity_category, so this association will never be surfaced as traversable when entity_category == "biolink:PhenotypicFeature". Today this looks harmless because that endpoint is only wired into PageGridPlayground.vue, not the live node pages — the actual "Treatments" section on Disease/Phenotype pages is driven by association_counts (Solr facet on real subject/object membership), which doesn't consult this field at all. But it's worth double-checking this is intentional (vs. just an oversight from generalizing to "Disease" as the object type), since the entity-grid feature seems to be actively under development and could start relying on this mapping soon.
• The PR description flags this itself, but worth restating: parseRetrievalSources assumes each sources entry may be a JSON-encoded array (nested chain) or a single object — this is inferred/tested against a hand-built fixture, not confirmed against a real API response yet. Since sources: Optional[list[str]] has no documented shape in the backend model, it'd be good to sanity-check on the deploy preview that MEDIC/CTD records actually serialize this way before merging out of draft.

Performance
No concerns — the new modal fields are simple computed properties over already-fetched association data, no extra requests introduced.

Security
No concerns — no new user input surfaces, and the JSON.parse in retrievalSources.ts is wrapped in try/catch so malformed data can't throw uncaught, and Vue's template interpolation escapes rendered values by default.

Test coverage

• Good frontend unit coverage for the new parseRetrievalSources helper (nested chain, empty, malformed).
• Backend changes are covered only via fixture updates for the label rename (per the PR description, existing association-counts/mapping tests pass) — there's no new backend test asserting the "Indications"/"Treatments" mapping actually resolves correctly for a Chemical/Disease/Phenotype entity, which would help catch the object_category gap above if it ever becomes load-bearing.
• As noted in the PR description, this hasn't been visually verified against a live backend yet (beta runs the old backend) — worth doing before taking it out of draft, especially for the nested provenance rendering.

Overall this looks solid and low-risk to merge once visually verified against the new backend. The object_category scoping is the only thing I'd flag as worth a second look, though it doesn't appear to affect current production behavior.

Superseded by #1357, which merges CTD in-app via the flexible section-matching enabler (#1355) instead of the single-category approach. Keeping the same modal-enrichment work (cherry-picked).