feat(peopledatalabs): add People Data Labs integration by waleedlatif1 · Pull Request #4513 · simstudioai/sim

waleedlatif1 · 2026-05-08T02:46:28Z

Summary

Adds the People Data Labs integration to Sim with 11 operations: person enrich/identify/search/bulk, company enrich/search/bulk/clean, location and school cleaners, and field autocomplete.
API-key auth via X-Api-Key. New block uses AuthMode.ApiKey, brand color #4831C3, and the official PDL icon.
Every tool was cross-validated against PDL's official API docs: scroll_token pagination (not from), top-level likelihood on company enrich, per-item likelihood on bulk company, full autocomplete field enum (location_name over deprecated location), correct dataset values, cleaners as POST, and 404-as-no-match handling.

Test plan

Person Enrich by email returns matched record + likelihood
Person Identify returns up to 20 candidate matches with match_score
Person Search via SQL returns results + scroll_token for pagination
Bulk Person Enrich processes a JSON array and echoes per-item metadata
Company Enrich by website returns top-level likelihood
Company Search returns total + scroll_token
Bulk Company Enrich returns per-item likelihood
Company / Location / School Cleaner POSTs return canonical records (or matched: false on 404)
Autocomplete returns suggestions with meta payload for company field

Add 11 PDL operations: person enrich/identify/search/bulk, company enrich/search/bulk/clean, location/school cleaners, and autocomplete. All endpoints, params, and response shapes verified against official PDL docs (scroll_token pagination, top-level likelihood on company enrich, per-item likelihood on bulk company, full autocomplete field enum). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

vercel · 2026-05-08T02:46:33Z

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment

Project	Deployment	Actions	Updated (UTC)
docs	Skipped		May 8, 2026 5:53pm

cursor · 2026-05-08T02:46:37Z

PR Summary

Medium Risk
Adds a new third-party integration with 11 API-backed tools and new parameter-mapping logic, which could affect request correctness and error handling. Risk is moderated by being additive and scoped to a new peopledatalabs type.

Overview
Adds a new People Data Labs integration end-to-end: new block configuration (API-key auth, operation selector, and parameter normalization to prevent stale UI values leaking across operations) plus registration in the block registry.

Introduces 11 new PDL tools (person/company enrich, search, identify, bulk enrich, cleaners, autocomplete) with typed request/response shaping, consistent X-Api-Key auth, and 404-as-no-match handling, and wires them into the global tool registry.

Updates docs and landing/integrations metadata to list the new tool, adds a PeopleDataLabsIcon, and maps the peopledatalabs type to that icon across docs and the integrations page.

^{Reviewed by Cursor Bugbot for commit 80f1279. Configure here.}

greptile-apps · 2026-05-08T02:50:01Z

Greptile Summary

This PR adds a complete People Data Labs integration with 11 operations spanning person and company enrichment, search, bulk processing, cleaner utilities, and field autocomplete. The implementation is thorough: API-key auth flows correctly through user-only visibility, 404 responses are handled as no-match rather than errors, scroll_token pagination is used correctly, and the multi-operation params remapping function carefully resets shared fields before repopulating them per-operation to prevent stale-value leakage — all hardened across several prior review rounds.

11 new tools registered in both the tool and block registries, each with typed params, transformResponse, and projected output schemas.
Multi-operation block with a careful alias-reset strategy in tools.config.params that prevents cross-operation field leakage confirmed clean after prior fixes.
Minor gaps remaining: the name input description is ambiguous (person name vs company name), and the cleaner tools send an empty POST body without error when no identifier field is provided.

Confidence Score: 5/5

Safe to merge — all tool-to-block parameter mappings are correct, auth is properly scoped, and the cross-operation field-reset logic is solid after prior iterations.

The integration is a clean additive change with no modifications to existing tools. All prior review findings have been addressed. The two remaining items are a misleading description on a dual-purpose input and missing pre-flight validation in cleaner tools — neither causes silent wrong results or data loss.

No files require special attention. The company_clean.ts and school_clean.ts body builders are worth a second look for the missing empty-input guard, but this is non-blocking.

Important Files Changed

Filename	Overview
apps/sim/blocks/blocks/peopledatalabs.ts	New block defining 11 PDL operations; params remapping is thorough after prior fixes.
apps/sim/tools/peopledatalabs/utils.ts	buildQueryString and projection helpers look correct.
apps/sim/tools/peopledatalabs/person_enrich.ts	Adds name param to both PdlPersonEnrichParams and buildQueryString call, addressing prior review findings.
apps/sim/tools/peopledatalabs/person_identify.ts	Full identify params forwarded correctly in query string.
apps/sim/tools/peopledatalabs/company_enrich.ts	Reads top-level likelihood from response as documented; correctly handles 404 as no-match.
apps/sim/tools/peopledatalabs/bulk_person_enrich.ts	JSON-parses and validates the requests array; maps per-item likelihood and metadata correctly.
apps/sim/tools/peopledatalabs/bulk_company_enrich.ts	Same pattern as bulk person enrich; correctly reads per-item likelihood at item level.
apps/sim/tools/peopledatalabs/company_clean.ts	POST body builder correctly only sends provided fields; 404 handled as no-match.
apps/sim/tools/peopledatalabs/types.ts	Comprehensive type definitions for all PDL request/response shapes.
apps/sim/tools/peopledatalabs/person_search.ts	Uses scroll_token for pagination correctly.
apps/sim/tools/peopledatalabs/company_search.ts	Same pattern as person search; correctly returns scroll_token for pagination.

_{Reviews (11): Last reviewed commit: "fix(peopledatalabs): isolate company `na..." | Re-trigger Greptile}

waleedlatif1 · 2026-05-08T02:50:47Z

@cursor review

waleedlatif1 · 2026-05-08T02:50:52Z

@greptileai

waleedlatif1 · 2026-05-08T02:51:51Z

@greptile

waleedlatif1 · 2026-05-08T02:51:52Z

@cursor review

…peration - min_likelihood now only shows for pdl_person_enrich (Person Identify ignores it) - ticker, pdl_id, company_location now only show for pdl_company_enrich (Company Cleaner only accepts name/website/profile) Addresses Greptile P1 review on PR #4513.

waleedlatif1 · 2026-05-08T02:59:08Z

@greptile

waleedlatif1 · 2026-05-08T02:59:09Z

@cursor review

Param renames (company_profile→profile, company_location→location, school_*→*, bulk_*_requests→requests, autocomplete_size→size, etc.) now run only when the matching operation is selected, and stale alternate-operation values are stripped from the request. This prevents values left over from a prior operation switch from leaking into the current API call (e.g. a company LinkedIn URL overwriting a person profile, or a stale search size overwriting autocomplete size). Addresses Cursor Bugbot review on PR #4513.

waleedlatif1 · 2026-05-08T03:06:19Z

@greptile

waleedlatif1 · 2026-05-08T03:06:20Z

@cursor review

The PeopleDataLabsIcon was hardcoded to white, leaving it invisible on light backgrounds when rendered outside its bgColor container (e.g., search results, menus, docs). Switch to currentColor so it inherits the surrounding text color. Addresses Cursor Bugbot review on PR #4513.

waleedlatif1 · 2026-05-08T03:16:35Z

@greptile

waleedlatif1 · 2026-05-08T03:16:36Z

@cursor review

cursor

✅ Bugbot reviewed your changes and found no new issues!

Comment @cursor review or bugbot run to trigger another review on this PR

^{Reviewed by Cursor Bugbot for commit f2a3c9f. Configure here.}

…te) per operation The block has subBlocks whose raw IDs collide with PDL API param names (profile, location for person; name, website for company). Their values persist across operation switches even though the UI hides them, so a person LinkedIn URL could leak into a Company Enrich request, etc. Reset these shared targets and repopulate them only from inputs that belong to the active operation. Addresses Greptile P1 review on PR #4513.

waleedlatif1 · 2026-05-08T04:09:56Z

@greptile

- person_identify: short-circuit on PDL 404 (no-match), matching the person_enrich pattern - company_search: drop unsupported `dataset` param (PDL company search docs do not list it) - block: expose `min_likelihood` for `pdl_company_enrich` (PDL Company Enrichment supports min_likelihood) - location_clean: surface `subregion`; drop phantom `latitude`/ `longitude` (PDL only returns `geo` as a "lat,lon" string) - school_clean: surface `domain` and `location_continent` from the nested `location` object - docs icon: switch fill to `currentColor` so the icon renders on light backgrounds

waleedlatif1 · 2026-05-08T04:46:58Z

@greptile

waleedlatif1 · 2026-05-08T04:47:01Z

@cursor review

The shared `name` reset at the top of `tools.config.params` was only repopulated for the company-side operations, so any programmatic `name` input to `pdl_person_enrich` or `pdl_person_identify` was silently dropped. Both PDL endpoints accept `name` as a full-name match parameter.

cursor

✅ Bugbot reviewed your changes and found no new issues!

Comment @cursor review or bugbot run to trigger another review on this PR

^{Reviewed by Cursor Bugbot for commit 092dfba. Configure here.}

waleedlatif1 · 2026-05-08T05:38:51Z

@greptile

waleedlatif1 · 2026-05-08T05:38:54Z

@cursor review

Add the `name` parameter to `PdlPersonEnrichParams`, the tool's params definition, and the URL builder. PDL Person Enrichment accepts `name` as a full-name match alternative to first_name + last_name; without it, programmatic `name` input was silently dropped before reaching the API.

waleedlatif1 · 2026-05-08T17:07:54Z

@greptile

waleedlatif1 · 2026-05-08T17:07:57Z

@cursor review

Rename Company Name subBlock id from `name` to `company_name` so a stale company value can't leak into Person Enrich/Identify when the user switches operations. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

waleedlatif1 · 2026-05-08T17:39:56Z

@greptile

waleedlatif1 · 2026-05-08T17:40:00Z

@cursor review

cursor

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 80f1279. Configure here.}

`pdl_clean_location` and `pdl_clean_school` were only restoring values from UI subBlock IDs (`clean_location_input`, `school_*`). Programmatic callers using the declared `location`/`name`/`website`/`profile` inputs had their values dropped after the shared-field reset. Add fallbacks so both UI and programmatic inputs flow through. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

…lete text - Company Enrich and Clean Company now fall back to programmatic `params.profile` / `params.location` when the UI-scoped `company_profile` / `company_location` are absent. Mirrors the fallback pattern already used for `name`. - Autocomplete `text` subBlock is now required when operation is autocomplete — PDL requires it for nearly all field values. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

vercel Bot deployed to Preview May 8, 2026 02:50 View deployment

greptile-apps Bot reviewed May 8, 2026

View reviewed changes

Comment thread apps/sim/blocks/blocks/peopledatalabs.ts

greptile-apps Bot reviewed May 8, 2026

View reviewed changes

Comment thread apps/sim/blocks/blocks/peopledatalabs.ts