Merged
Conversation
Contributor
|
Caution Review failedPull request was closed or merged during review 📝 WalkthroughWalkthroughAdds comprehensive link type and record linking infrastructure to the CRM system, including mutations for creating/deactivating link types and linking records with cardinality enforcement, queries for retrieving linked records grouped by link type, extensive validation for organization isolation and active-link constraints, and full test coverage for all CRUD operations. Changes
Sequence Diagram(s)sequenceDiagram
participant Client
participant Backend as Convex Backend
participant DB as Database
participant Audit as Audit Log
Client->>Backend: createLinkType(name, sourceObjDefId, targetObjDefId, cardinality)
Backend->>Backend: Validate ctx.viewer.orgId present
Backend->>DB: Load sourceObjectDef
Backend->>DB: Load targetObjectDef
Backend->>Backend: Verify both exist, active, belong to org
Backend->>DB: Insert linkTypeDefs record (isActive: true)
Backend->>Audit: Log crm.linkType.created
Backend->>Client: Return linkTypeDefId
Client->>Backend: deactivateLinkType(linkTypeDefId)
Backend->>DB: Load linkTypeDef, verify org ownership
Backend->>DB: Query recordLinks by linkTypeDefId (isDeleted: false)
Backend->>Backend: Throw if active links exist
DB->>DB: (No active links found)
Backend->>DB: Patch linkTypeDef (isActive: false)
Backend->>Audit: Log crm.linkType.deactivated
Backend->>Client: Return success
sequenceDiagram
participant Client
participant Backend as Convex Backend
participant DB as Database
participant Audit as Audit Log
Client->>Backend: createLink(linkTypeDefId, sourceKind, sourceId, targetKind, targetId)
Backend->>Backend: Validate ctx.viewer.orgId present
Backend->>DB: Load linkTypeDef, verify active & org ownership
Backend->>DB: Validate source entity exists & org-owned
Backend->>DB: Validate target entity exists & org-owned
Backend->>DB: Verify object types match linkTypeDef expectations
Backend->>Backend: Check cross-org pairing (reject if mismatch)
Backend->>DB: Query existing active links (by_org_source index)
Backend->>Backend: Reject if duplicate in forward or reverse direction
Backend->>Backend: Enforce cardinality constraints (one_to_one/one_to_many/many_to_many)
DB->>DB: (Cardinality checks pass)
Backend->>DB: Insert recordLinks record (isDeleted: false)
Backend->>Audit: Log crm.link.created
Backend->>Client: Return linkId
Client->>Backend: deleteLink(linkId)
Backend->>DB: Load recordLinks, verify org ownership & not already deleted
Backend->>DB: Patch isDeleted to true
Backend->>Audit: Log crm.link.deleted
Backend->>Client: Return success
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Possibly related issues
Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 1 | ❌ 2❌ Failed checks (1 warning, 1 inconclusive)
✅ Passed checks (1 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Owner
Author
This stack of pull requests is managed by Graphite. Learn more about stacking. |
Reviewer's GuideImplements polymorphic CRM record linking, including admin CRUD for link types, data-plane link create/delete mutations, bidirectional link queries grouped by link type, and a small date-handling bugfix plus spec updates for ENG-254 and ENG-257. Sequence diagram for createLink CRM mutationsequenceDiagram
actor User
participant Client
participant createLink_handler
participant MutationCtx
participant ConvexDb as ConvexDb_recordLinks
participant auditLog_module
User->>Client: Request createLink(linkTypeDefId, sourceKind, sourceId, targetKind, targetId)
Client->>createLink_handler: crmMutation.createLink(args)
createLink_handler->>MutationCtx: Read viewer.orgId
alt orgId is missing
createLink_handler-->>Client: ConvexError Org context required
else orgId present
createLink_handler->>ConvexDb: get(linkTypeDefId)
ConvexDb-->>createLink_handler: linkTypeDef
createLink_handler->>ConvexDb: get(sourceObjectDefId)
ConvexDb-->>createLink_handler: sourceObjectDef
createLink_handler->>ConvexDb: get(targetObjectDefId)
ConvexDb-->>createLink_handler: targetObjectDef
createLink_handler->>createLink_handler: validateEntityExists(source)
createLink_handler->>createLink_handler: validateEntityExists(target)
createLink_handler->>ConvexDb: query(recordLinks by_org_source)
ConvexDb-->>createLink_handler: existingSourceLinks
createLink_handler->>createLink_handler: check duplicate and cardinality
createLink_handler->>ConvexDb: insert(recordLinks, newLink)
ConvexDb-->>createLink_handler: linkId
createLink_handler->>auditLog_module: log crm.link.created
auditLog_module-->>createLink_handler: ack
createLink_handler-->>Client: linkId
end
Client-->>User: Link created
Sequence diagram for getLinkedRecords CRM querysequenceDiagram
actor User
participant Client
participant getLinkedRecords_handler
participant QueryCtx
participant ConvexDb as ConvexDb_recordLinks
User->>Client: View linked records for entity
Client->>getLinkedRecords_handler: crmQuery.getLinkedRecords(recordId, recordKind, direction)
getLinkedRecords_handler->>QueryCtx: Read viewer.orgId
alt orgId is missing
getLinkedRecords_handler-->>Client: ConvexError Org context required
else orgId present
getLinkedRecords_handler->>getLinkedRecords_handler: Resolve direction flags
opt outbound requested
getLinkedRecords_handler->>ConvexDb: query(recordLinks by_org_source)
ConvexDb-->>getLinkedRecords_handler: outboundLinks
end
opt inbound requested
getLinkedRecords_handler->>ConvexDb: query(recordLinks by_org_target)
ConvexDb-->>getLinkedRecords_handler: inboundLinks
end
getLinkedRecords_handler->>getLinkedRecords_handler: Group links by linkTypeDefId
getLinkedRecords_handler->>ConvexDb: get(linkTypeDefIds)
ConvexDb-->>getLinkedRecords_handler: linkTypeDefs
loop each outbound group
getLinkedRecords_handler->>ConvexDb: get(peer records for outbound)
ConvexDb-->>getLinkedRecords_handler: records with labelValue
getLinkedRecords_handler->>getLinkedRecords_handler: build LinkedRecord for outbound
end
loop each inbound group
getLinkedRecords_handler->>ConvexDb: get(peer records for inbound)
ConvexDb-->>getLinkedRecords_handler: records with labelValue
getLinkedRecords_handler->>getLinkedRecords_handler: build LinkedRecord for inbound
end
getLinkedRecords_handler-->>Client: LinkGroup[]
end
Client-->>User: Render grouped linked records
ER diagram for CRM record link data modelerDiagram
objectDefs {
string _id
string orgId
string name
string nativeTable
boolean isActive
}
linkTypeDefs {
string _id
string orgId
string name
string sourceObjectDefId
string targetObjectDefId
string cardinality
boolean isActive
number createdAt
}
recordLinks {
string _id
string orgId
string linkTypeDefId
string sourceObjectDefId
string sourceKind
string sourceId
string targetObjectDefId
string targetKind
string targetId
boolean isDeleted
number createdAt
string createdBy
}
records {
string _id
string orgId
string objectDefId
boolean isDeleted
string labelValue
}
mortgages {
string _id
string orgId
}
borrowers {
string _id
string orgId
}
lenders {
string _id
string orgId
}
brokers {
string _id
string orgId
}
deals {
string _id
string orgId
}
obligations {
string _id
string orgId
}
objectDefs ||--o{ linkTypeDefs : sourceObject
objectDefs ||--o{ linkTypeDefs : targetObject
linkTypeDefs ||--o{ recordLinks : typedBy
objectDefs ||--o{ records : defines
records ||--o{ recordLinks : canBeSource
records ||--o{ recordLinks : canBeTarget
objectDefs ||--o{ mortgages : native
objectDefs ||--o{ borrowers : native
objectDefs ||--o{ lenders : native
objectDefs ||--o{ brokers : native
objectDefs ||--o{ deals : native
objectDefs ||--o{ obligations : native
mortgages ||--o{ recordLinks : canBeLinked
borrowers ||--o{ recordLinks : canBeLinked
lenders ||--o{ recordLinks : canBeLinked
brokers ||--o{ recordLinks : canBeLinked
deals ||--o{ recordLinks : canBeLinked
obligations ||--o{ recordLinks : canBeLinked
Class diagram for CRM record linking modulesclassDiagram
class recordLinks_module {
+getNativeEntity(ctx, tableName, entityId) Promise~object~
+validateEntityExists(ctx, kind, entityId, orgId, objectDef) Promise~void~
+createLink(ctx, args) Promise~Id_recordLinks~
+deleteLink(ctx, args) Promise~void~
}
class linkQueries_module {
+resolveLinkedRecords(ctx, links, direction) Promise~LinkedRecord[]~
+getLinkedRecords(ctx, args) Promise~LinkGroup[]~
+getLinkTypesForObject(ctx, args) Promise~Doc_linkTypeDefs[]~
}
class linkTypes_module {
+createLinkType(ctx, args) Promise~Id_linkTypeDefs~
+deactivateLinkType(ctx, args) Promise~void~
+listLinkTypes(ctx) Promise~Doc_linkTypeDefs[]~
}
class LinkedRecord {
+string linkId
+Id_linkTypeDefs linkTypeDefId
+string recordId
+string recordKind
+Id_objectDefs objectDefId
+string labelValue
}
class LinkGroup {
+string direction
+LinkedRecord[] links
+Id_linkTypeDefs linkTypeDefId
+string linkTypeName
}
class auditLog_module {
+log(ctx, entry) Promise~void~
}
class crmMutation {
+input(schema)
+handler(fn)
+public()
}
class crmAdminMutation {
+input(schema)
+handler(fn)
+public()
}
class crmQuery {
+input(schema)
+handler(fn)
+public()
}
class crmAdminQuery {
+input(schema)
+handler(fn)
+public()
}
class cardinalityValidator {
}
class ConvexDb {
+get(id)
+insert(table, doc)
+patch(id, fields)
+query(table)
}
class MutationCtx {
+ConvexDb db
+Viewer viewer
}
class QueryCtx {
+ConvexDb db
+Viewer viewer
}
class Viewer {
+string orgId
+string authId
}
recordLinks_module --> MutationCtx
recordLinks_module --> auditLog_module
recordLinks_module --> crmMutation
linkQueries_module --> QueryCtx
linkQueries_module --> LinkedRecord
linkQueries_module --> LinkGroup
linkQueries_module --> crmQuery
linkTypes_module --> MutationCtx
linkTypes_module --> auditLog_module
linkTypes_module --> crmAdminMutation
linkTypes_module --> crmAdminQuery
linkTypes_module --> cardinalityValidator
MutationCtx --> ConvexDb
QueryCtx --> ConvexDb
MutationCtx --> Viewer
QueryCtx --> Viewer
File-Level Changes
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
![sourcery-ai[bot]](https://avatars.githubusercontent.com/in/48477?s=60&v=4){kind=small}
There was a problem hiding this comment.
Sorry @Connorbelez, you have reached your weekly rate limit of 500000 diff characters.
Please try again later or upgrade to continue using Sourcery
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
There was a problem hiding this comment.
Your free trial has ended. If you'd like to continue receiving code reviews, you can add a payment method here.
There was a problem hiding this comment.
Pull request overview
Implements ENG-257 by introducing polymorphic CRM record linking (admin-managed link types, data-plane link create/delete, and bidirectional link queries) and updates date filter encoding to be UTC-safe, alongside supporting schema/type changes and specs/tests.
Changes:
- Add Convex CRUD for
linkTypeDefs(create, deactivate, list) and record linking mutations (create + soft-delete) with audit logging. - Add bidirectional link queries (
getLinkedRecords,getLinkTypesForObject) and supporting schema/index + type updates. - Harden admin FilterBuilder date encoding to parse
yyyy-mm-ddas UTC, and add ENG-257/ENG-254 spec documentation + new tests.
Reviewed changes
Copilot reviewed 19 out of 19 changed files in this pull request and generated 6 comments.
Show a summary per file
| File | Description |
|---|---|
| src/components/admin/shell/FilterBuilder.tsx | Adds explicit UTC parsing for date filter encoding to avoid timezone inconsistencies. |
| specs/ENG-257/tasks.md | Adds master task list for ENG-257 implementation chunks. |
| specs/ENG-257/chunks/manifest.md | Adds chunk manifest for ENG-257 spec breakdown. |
| specs/ENG-257/chunks/chunk-03-bidirectional-queries/tasks.md | Documents requirements for bidirectional link queries and verification steps. |
| specs/ENG-257/chunks/chunk-03-bidirectional-queries/context.md | Provides schema/type/context references for implementing bidirectional queries. |
| specs/ENG-257/chunks/chunk-02-record-linking/tasks.md | Documents requirements for record linking mutations and validation rules. |
| specs/ENG-257/chunks/chunk-02-record-linking/context.md | Provides schema + validation/audit patterns for linking mutations. |
| specs/ENG-257/chunks/chunk-01-link-type-crud/tasks.md | Documents requirements for admin link type CRUD. |
| specs/ENG-257/chunks/chunk-01-link-type-crud/context.md | Provides schema + middleware/audit patterns for link type CRUD. |
| specs/ENG-254/chunks/chunk-01-calendar-query/context.md | Updates calendar query spec with truncation behavior and contract fields. |
| convex/schema.ts | Adds org-scoped indexes for link types; standardizes recordLink kind validation via shared validator. |
| convex/crm/validators.ts | Introduces entityKindValidator for consistent `"record" |
| convex/crm/types.ts | Tightens LinkedRecord.linkId type to Id<"recordLinks">. |
| convex/crm/recordQueries.ts | Aligns getRecord link output with updated LinkedRecord.linkId type. |
| convex/crm/recordLinks.ts | Adds createLink/deleteLink mutations plus entity validation helpers and audit logging. |
| convex/crm/linkTypes.ts | Adds admin mutations/queries for link type creation, deactivation, and listing. |
| convex/crm/linkQueries.ts | Adds bidirectional link queries grouped by link type and direction. |
| convex/crm/tests/recordLinks.test.ts | Adds mutation tests for cardinality, duplicates, org isolation, and soft-delete behaviors. |
| convex/crm/tests/linkTypes.test.ts | Adds tests for link type creation, listing, deactivation, and org isolation. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 3
🧹 Nitpick comments (4)
specs/ENG-257/chunks/chunk-03-bidirectional-queries/tasks.md (1)
17-24: Spec references non-org-scoped indexes, but implementation correctly uses org-scoped variants.The spec states
getLinkTypesForObjectshould query viaby_source_objectandby_target_objectindexes, but the implementation inconvex/crm/linkQueries.ts(lines 221-232) correctly uses the org-scopedby_org_source_objectandby_org_target_objectindexes instead. The implementation is preferable for org isolation. Consider updating the spec to match.📝 Suggested spec update
## T-007: Create `getLinkTypesForObject` query - **File:** `convex/crm/linkQueries.ts` (add to same file) - **Details:** - Use `crmQuery` - Args: `{ objectDefId: v.id("objectDefs") }` - - Query `linkTypeDefs` via `by_source_object` and `by_target_object` indexes + - Query `linkTypeDefs` via `by_org_source_object` and `by_org_target_object` indexes - Filter `isActive === true` - Return combined list (deduplicated by _id) — used by ENG-258's "Add Link" dialog🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@specs/ENG-257/chunks/chunk-03-bidirectional-queries/tasks.md` around lines 17 - 24, Update the spec for getLinkTypesForObject to match the implementation: state that the query (crmQuery getLinkTypesForObject) should use the org-scoped indexes by_org_source_object and by_org_target_object (not the non-org-scoped by_source_object/by_target_object), accept args { objectDefId: v.id("objectDefs") }, filter linkTypeDefs where isActive === true, and return a combined deduplicated list by _id for use by the "Add Link" dialog.specs/ENG-257/chunks/chunk-03-bidirectional-queries/context.md (1)
32-42: Minor documentation drift:linkIdtype shown asstringbut implementation usesId<"recordLinks">.The
LinkedRecordinterface in this context doc showslinkId: string, but the actual implementation inconvex/crm/linkQueries.ts(line 45) assignslinkId: link._idwhich is typed asId<"recordLinks">. The type inference is correct in the implementation; consider updating the spec to reflect the actual type for accuracy.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@specs/ENG-257/chunks/chunk-03-bidirectional-queries/context.md` around lines 32 - 42, The spec's LinkedRecord type is out of sync: change the linkId field from string to the proper Id<"recordLinks"> type to match the implementation (see linkQueries.ts where linkId is set from link._id) and ensure the context doc mirrors the type defined in convex/crm/types.ts for LinkedRecord; update the type annotation in specs/ENG-257/chunks/chunk-03-bidirectional-queries/context.md accordingly.convex/crm/linkTypes.ts (1)
125-137: Consider addingisActiveto theby_orgindex for query efficiency.The current implementation fetches all
linkTypeDefsfor the org, then filtersisActivein memory. While acceptable for admin operations with typically few link types, adding a composite index likeby_org_active: ["orgId", "isActive"]would enable more efficient querying as usage scales.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@convex/crm/linkTypes.ts` around lines 125 - 137, The current listLinkTypes handler queries linkTypeDefs using the by_org index then filters isActive in memory; add a composite index (e.g., by_org_active with keys ["orgId", "isActive"]) and update the query in listLinkTypes to use withIndex("by_org_active", q => q.eq("orgId", orgId).eq("isActive", true)) so the database returns only active link types instead of fetching all and filtering; ensure the index name matches your schema and tests are updated where linkTypeDefs/by_org is referenced.convex/crm/recordLinks.ts (1)
168-186: Tighten these lookups to the active link type.Both queries load all links for the source/target and then filter in memory for
linkTypeDefIdandisDeleted. That makescreateLinkcost grow with total fanout on each side, including soft-deleted history, even though the handler only needs a few existence checks. Convex calls out read/write limits at the mutation level and recommends indexing queries to reduce scanned documents, so this is worth narrowing before link volume grows. (docs.convex.dev)Also applies to: 189-249
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@convex/crm/recordLinks.ts` around lines 168 - 186, The queries that populate existingSourceLinks and existingTargetLinks (the .withIndex("by_org_source", ...) and .withIndex("by_org_target", ...) calls) currently fetch all links for the source/target and then filter in memory; update those index queries to include eq("linkTypeDefId", args.linkTypeDefId) and eq("isDeleted", false) so the database only scans active links of the requested type, and apply the same tightening to the other similar queries in this file (the subsequent queries around the createLink logic referenced in lines 189-249) to avoid scanning soft-deleted history and reduce mutation read cost.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@convex/crm/recordLinks.ts`:
- Around line 24-45: The getNativeEntity function is casting plain strings to
Convex Ids before fetching, which is unsafe; replace the direct cast +
ctx.db.get calls with validation via ctx.db.normalizeId(tableName, entityId) and
if normalizeId returns null, return null (do not call ctx.db.get); otherwise
pass the normalized Id value to ctx.db.get. Apply the same pattern to other
places handling unvalidated polymorphic IDs (e.g., any code working with
sourceId/targetId in this file between the earlier and later switch blocks) so
every ctx.db.get uses a normalized Id from ctx.db.normalizeId(tableName, id)
rather than a raw cast.
In `@specs/ENG-254/chunks/chunk-01-calendar-query/context.md`:
- Around line 50-51: The spec's type for skippedFilters is wrong: the
implementation's CalendarData contract uses a numeric skipped count, not
string[], so update the spec to match the code by changing skippedFilters:
string[] to skippedFilters: number (or rename to skippedCount: number if
aligning naming with CalendarData); ensure references to
skippedFilters/CalendarData in the document and any examples reflect the numeric
type so clients correctly expect a count rather than an array.
In `@src/components/admin/shell/FilterBuilder.tsx`:
- Around line 38-45: The parseDateToUtcMs function currently uses Date.UTC which
normalizes invalid calendar dates; update parseDateToUtcMs to reject impossible
dates by validating the produced UTC date matches the input components: after
parsing year/month/day and doing basic range checks, compute ms = Date.UTC(year,
month - 1, day) then create a Date from ms and compare getUTCFullYear(),
getUTCMonth()+1 and getUTCDate() against the original year, month and day; if
any differ return null, otherwise return ms. This ensures functions like
parseDateToUtcMs only accept true calendar dates.
---
Nitpick comments:
In `@convex/crm/linkTypes.ts`:
- Around line 125-137: The current listLinkTypes handler queries linkTypeDefs
using the by_org index then filters isActive in memory; add a composite index
(e.g., by_org_active with keys ["orgId", "isActive"]) and update the query in
listLinkTypes to use withIndex("by_org_active", q => q.eq("orgId",
orgId).eq("isActive", true)) so the database returns only active link types
instead of fetching all and filtering; ensure the index name matches your schema
and tests are updated where linkTypeDefs/by_org is referenced.
In `@convex/crm/recordLinks.ts`:
- Around line 168-186: The queries that populate existingSourceLinks and
existingTargetLinks (the .withIndex("by_org_source", ...) and
.withIndex("by_org_target", ...) calls) currently fetch all links for the
source/target and then filter in memory; update those index queries to include
eq("linkTypeDefId", args.linkTypeDefId) and eq("isDeleted", false) so the
database only scans active links of the requested type, and apply the same
tightening to the other similar queries in this file (the subsequent queries
around the createLink logic referenced in lines 189-249) to avoid scanning
soft-deleted history and reduce mutation read cost.
In `@specs/ENG-257/chunks/chunk-03-bidirectional-queries/context.md`:
- Around line 32-42: The spec's LinkedRecord type is out of sync: change the
linkId field from string to the proper Id<"recordLinks"> type to match the
implementation (see linkQueries.ts where linkId is set from link._id) and ensure
the context doc mirrors the type defined in convex/crm/types.ts for
LinkedRecord; update the type annotation in
specs/ENG-257/chunks/chunk-03-bidirectional-queries/context.md accordingly.
In `@specs/ENG-257/chunks/chunk-03-bidirectional-queries/tasks.md`:
- Around line 17-24: Update the spec for getLinkTypesForObject to match the
implementation: state that the query (crmQuery getLinkTypesForObject) should use
the org-scoped indexes by_org_source_object and by_org_target_object (not the
non-org-scoped by_source_object/by_target_object), accept args { objectDefId:
v.id("objectDefs") }, filter linkTypeDefs where isActive === true, and return a
combined deduplicated list by _id for use by the "Add Link" dialog.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: df02a803-d176-4dc5-858a-6a05b0d52562
📒 Files selected for processing (19)
convex/crm/__tests__/linkTypes.test.tsconvex/crm/__tests__/recordLinks.test.tsconvex/crm/linkQueries.tsconvex/crm/linkTypes.tsconvex/crm/recordLinks.tsconvex/crm/recordQueries.tsconvex/crm/types.tsconvex/crm/validators.tsconvex/schema.tsspecs/ENG-254/chunks/chunk-01-calendar-query/context.mdspecs/ENG-257/chunks/chunk-01-link-type-crud/context.mdspecs/ENG-257/chunks/chunk-01-link-type-crud/tasks.mdspecs/ENG-257/chunks/chunk-02-record-linking/context.mdspecs/ENG-257/chunks/chunk-02-record-linking/tasks.mdspecs/ENG-257/chunks/chunk-03-bidirectional-queries/context.mdspecs/ENG-257/chunks/chunk-03-bidirectional-queries/tasks.mdspecs/ENG-257/chunks/manifest.mdspecs/ENG-257/tasks.mdsrc/components/admin/shell/FilterBuilder.tsx
… dates like Feb 31
…ilters, and ID validation in recordLinks - Use ctx.db.normalizeId() instead of unsafe as Id<> casts in getNativeEntity and validateEntityExists - Add isDeleted=false and linkTypeDefId filters to source/target link queries to reduce reads - Fix reverse duplicate check to actually query the reverse direction (B→A when A→B exists) - Enforce full one_to_one semantics by checking cross-direction participation
Merged
Connorbelez
added a commit
that referenced
this pull request
Apr 20, 2026
eng-257 eng-257 ## Summary by Sourcery Introduce polymorphic CRM record linking with admin-managed link types and bidirectional queries, and harden date filter handling to be timezone-safe. New Features: - Add admin CRUD operations for link type definitions, including creation, deactivation, and listing of link types scoped to an org. - Implement record linking mutations that create and soft-delete polymorphic links between CRM records and native entities with cardinality and org-ownership enforcement. - Expose queries to fetch linked records bidirectionally, grouped by link type, and to list applicable link types for a given object definition. Bug Fixes: - Change date filter encoding to use explicit UTC parsing, avoiding cross-browser timezone inconsistencies when converting date strings to timestamps. Documentation: - Add ENG-257 specification chunks covering link type CRUD, record linking mutations, and bidirectional link queries, plus a master task list and chunk manifest. - Update ENG-254 calendar query spec to document result truncation behavior and extend the CalendarData contract with skippedFilters and truncated flags. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit ## Release Notes * **New Features** * Link type management: create, deactivate, and list link types for your organization * Record linking: establish and remove relationships between records * Bidirectional record queries: view all linked records and available link types for objects * Enhanced date validation in filter inputs with stricter parsing and error handling * **Tests** * Added comprehensive test coverage for link type creation, deactivation, and record linking operations with cardinality enforcement and organization isolation validation <!-- end of auto-generated comment: release notes by coderabbit.ai -->
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
eng-257
eng-257
Summary by Sourcery
Introduce polymorphic CRM record linking with admin-managed link types and bidirectional queries, and harden date filter handling to be timezone-safe.
New Features:
Bug Fixes:
Documentation:
Summary by CodeRabbit
Release Notes
New Features
Tests