Power your chat with B2B data: lead lists, company research, contact enrichment, and outreach personalization.

GitHub stars
298

Install

/plugin install vpai@vibeprospecting

Run it inside Claude Code after adding the marketplace with /plugin marketplace add explorium-ai/vibeprospecting-plugin.

Contents

Discovered from source

Skills 2

Vibe Prospecting → HubSpot field mapping

Vibe Prospecting prospects map to HubSpot Contacts. Vibe Prospecting businesses map to HubSpot Companies.

Vibe Prospecting field names are the column names in Vibe Prospecting results. HubSpot properties are shown with their label and internal name.

When this mapping applies. Whenever Vibe-sourced fields are written to HubSpot (insert or update), use the destinations in these tables, even if the user did not ask about mapping. This rule only decides where Vibe-sourced fields go. It does not authorize a write, and it does not add records or fields beyond what the user asked for.

Property definitions are the user's to manage

  • Never create or modify HubSpot property definitions or enumeration options.
  • Before writing, check that each mapped destination exists on the correct object (Contacts for prospects, Companies for businesses).
  • For ordinary mapped fields, if a destination is missing or cannot accept the raw value (for example an invalid enumeration option), omit that field, report its HubSpot label and internal name, and continue with the other mapped fields. The list name below is required by default; skip it only when the user explicitly requests that exception.
  • Do not silently substitute another destination, and do not transform, split, truncate or derive values to make them fit. The only scoped exception is the additional-phone extraction described below; do not apply it to other fields.

Custom destinations are marked (custom) and must be created by the user in HubSpot before they can receive data. They use the VP label prefix and vp_ internal-name prefix. Users may change these defaults in their own skill instance.

To change a destination: create the property yourself in HubSpot (Settings → Properties), or pick an existing one. Then edit the row in this file: update the HubSpot label, internal name, type and the custom marker. Keep the Vibe Prospecting field name unchanged. Reinstall the updated skill.

If you edit a destination in a mapping table, use that row's edited label, internal name, and type—not the default property names mentioned elsewhere in this file.

List name before export

By default, before each new export for a HubSpot push, suggest a list name based on the user's request and conversation, then ask the user to approve it or provide a different name. Do not offer the dataset name as a default. If the user already explicitly supplied a name for this export, that is their confirmation; do not ask again. Do not proceed with an unconfirmed or blank name, or silently reuse one from a different export.

The default destination on Contacts and Companies is VP List Name (vp_list_name, single-line text, custom). The user may explicitly choose a different existing destination for this push or explicitly ask to skip the list-name field. Honor that choice for this push without requiring a skill edit or reinstall. Do not infer permission to skip or change the destination from a missing property or general export approval. For a lasting change to the default, the user may edit this destination in the skill and reinstall it.

Unless the user explicitly asked to skip the list name, check the chosen destination's live HubSpot property definition on every target object type before export. It must exist and accept the confirmed name. If missing or incompatible, identify its label, internal name, and object type, and ask the user to create/fix it manually, choose another existing property, or explicitly request skipping the list-name field. Stop the push before export until a usable destination is verified or the user explicitly requests skipping it. Never create properties or silently substitute a destination.

When the list name is not skipped, write the confirmed name unchanged on every contact and company inserted or updated. Recheck the chosen destination before record writes; if it is no longer usable, stop until resolved or the user explicitly requests skipping it. If the user explicitly requests skipping it, omit the field and reflect that choice in the pre-write report. The name approval or skip request is not approval for export cost or HubSpot writes; retain the existing approval requirements. This is a field only: do not create a HubSpot list or add list memberships. Other fields retain ordinary mapping/reporting behavior without dedicated missing-field prompts.

Show the field mapping before writing to HubSpot (required)

After the Vibe Prospecting data is exported and before any record is created or updated in HubSpot, always show the user a field-mapping report for this specific push. Build it from a live check of the target HubSpot account, not from this file alone:

  1. Check every destination in the relevant table(s) below against the account. For enumeration destinations such as Lead Status, also confirm the required option exists.
  2. Check the exported data: note which Vibe fields actually have values for the rows being pushed.
  3. Present one table per object type being written, with these columns:
Vibe Prospecting fieldHubSpot labelInternal nameIn HubSpot?Written in this push?
  • In HubSpot?: ✅ Exists (mark standard vs custom where useful) or ❌ Missing.
  • Written in this push?: ✅ Yes; ➖ No data (destination exists but the field is empty for these rows, with a short reason, e.g. "emails only were requested"); or ❌ Skipped (destination missing or cannot accept the value).
  • Include every mapped field from the table, not only the ones being written, so the user sees what exists and what does not.
  • One report row per mapping-table row; never merge rows (e.g. VP Additional Phone Number 1–3 are three rows).
  • Put the list-name row first in each contact/company report table and bold its label and value so it stands out from other fields. Show the chosen destination or that the user explicitly requested skipping it. Unless explicitly skipped, stop if it cannot be written. Add rows for Lead Status on new contacts and associations being created (e.g. contact → existing company, with its ID).
  1. Then ask for approval of the write as usual. The mapping report does not replace the HubSpot record-level confirmation; show both, or combine them in one message.

Contacts: Vibe Prospecting prospect → HubSpot Contact

Vibe Prospecting fieldHubSpot labelHubSpot internal nameType
prospect_first_nameFirst NamefirstnameSingle-line text
prospect_last_nameLast NamelastnameSingle-line text
contact_professional_emailEmailemailSingle-line text
contact_mobile_phoneMobile Phone NumbermobilephoneSingle-line text
contact_phone_numbers (additional number 1; see extraction below)VP Additional Phone Number 1 (custom)vp_additional_phone_number_1Single-line text
contact_phone_numbers (additional number 2; see extraction below)VP Additional Phone Number 2 (custom)vp_additional_phone_number_2Single-line text
contact_phone_numbers (additional number 3; see extraction below)VP Additional Phone Number 3 (custom)vp_additional_phone_number_3Single-line text
prospect_job_titleJob TitlejobtitleSingle-line text
prospect_company_nameCompany NamecompanySingle-line text
prospect_company_websiteWebsite URLwebsiteSingle-line text
prospect_cityCitycitySingle-line text
prospect_region_nameState/RegionstateSingle-line text
prospect_country_nameCountry/RegioncountrySingle-line text
prospect_linkedinLinkedIn URLhs_linkedin_urlSingle-line text
prospect_idVP Prospect ID (custom)vp_prospect_idSingle-line text
business_idVP Business ID (custom)vp_business_idSingle-line text
prospect_full_nameVP Full Name (custom)vp_full_nameSingle-line text
prospect_job_departmentVP Job Department (custom)vp_job_departmentSingle-line text
prospect_job_seniority_levelVP Job Seniority Level (custom)vp_job_seniority_levelSingle-line text
contact_professional_email_statusVP Email Status (custom)vp_email_statusSingle-line text
prospect_experienceVP Experience (custom)vp_experienceMulti-line text
prospect_skillsVP Skills (custom)vp_skillsMulti-line text
prospect_interestsVP Interests (custom)vp_interestsMulti-line text

Lead Status. For contacts newly created from Vibe Prospecting data, set Lead Status (hs_lead_status) to the option labelled New, internal value NEW. Never change Lead Status on existing contacts. If hs_lead_status in the target account has no option labelled New, skip it and report; do not create or substitute an option.

Additional phone extraction

Keep contact_mobile_phone → mobilephone unchanged. Extra numbers come from the separate contact_phone_numbers export column, not from splitting contact_mobile_phone.

In the VP CSV export, contact_phone_numbers is a JSON-encoded array of phone-number strings, not an array of objects. Read the CSV with a proper CSV parser first, then JSON-decode the cell once; CSV quote escaping is not part of the JSON value. If the data reader already decoded the cell, use the array of strings directly. Walk its string items in array order. Do not split on commas, inspect object values, stringify other types, or extract numbers from arbitrary text. An empty cell is ordinary no data. A malformed JSON cell, non-array value (including a single plain-text phone number), or non-string array item leaves the additional-phone fields unmapped for that row rather than guessing a format.

For comparison only, trim surrounding whitespace and remove spaces, parentheses, hyphens, and dots from each number; preserve a leading + and any explicit country code. Exclude every value with the same comparison key as contact_mobile_phone, skip blank values, and deduplicate the remaining values by that key in first-occurrence order. The primary can occur anywhere in the array or be absent: do not drop the first item by position or require the primary to be present. Do not infer a country code, convert 00 prefixes, or drop extensions to force equality. Write each retained value exactly as supplied, not its comparison key.

Map the first three retained values to VP Additional Phone Number 1, 2, and 3 respectively. The limit is three additional numbers, beyond the primary. Omit unused destinations without clearing existing values; ignore excess values. A missing destination keeps its assigned slot: do not shift its number into another property. Phone fields use the same pre-write mapping report and skip/no-data handling as ordinary fields—no extra prompts, missing-phone warnings, or overflow counts.

The array may not be included in a particular export. If contact_phone_numbers is absent or empty, handle it as ordinary unavailable source data and preserve the primary mapping. Never invent extra numbers or fetch a different data source just to fill these fields.

Companies: Vibe Prospecting business → HubSpot Company

Vibe Prospecting fieldHubSpot labelHubSpot internal nameType
business_nameCompany namenameSingle-line text
business_domainCompany Domain NamedomainSingle-line text
business_websiteWebsite URLwebsiteSingle-line text
business_city_nameCitycitySingle-line text
business_regionState/RegionstateSingle-line text
business_country_nameCountry/RegioncountrySingle-line text
business_business_descriptionDescriptiondescriptionMulti-line text
prospect_company_linkedinLinkedIn Company Pagelinkedin_company_pageSingle-line text
business_logoLogo URLhs_logo_urlText (string)
business_number_of_employees_rangeEmployee rangehs_employee_rangeText (string)
business_yearly_revenue_rangeRevenue rangehs_revenue_rangeText (string)
business_idVP Business ID (custom)vp_business_idSingle-line text
business_naicsVP NAICS Code (custom)vp_naics_codeSingle-line text
business_naics_descriptionVP NAICS Description (custom)vp_naics_descriptionSingle-line text
business_sic_codeVP SIC Code (custom)vp_sic_codeSingle-line text
business_sic_code_descriptionVP SIC Code Description (custom)vp_sic_code_descriptionSingle-line text
business_business_intent_topicsVP Intent Topics (custom)vp_intent_topicsMulti-line text

Employee and revenue ranges go only to the range properties above, as the raw range text. Never write them to the numeric numberofemployees or annualrevenue properties.

row_num and created_at are export metadata and are not mapped.

Vibe Prospecting

B2B prospecting — companies, contacts, enrichment, events.

Platform

Before any prospecting work, detect your host using the rules below, read exactly one platform guide, and follow it for auth, workflow, flags, and troubleshooting.

Detecting your runtime (mandatory, in order)

  1. Host identity in your system context (check this first — overrides tool names):
    • Claude Code — context explicitly says you are Claude Code / Anthropic's CLI. → read claude-code.md only. Do not read claude-chat.md or cowork.md. MCP tools in the list (mcp__*__fetch-entities) do not change this.
    • OpenAI Codex — context identifies Codex or CODEX_SHELL=1. → codex.md
    • OpenClaw — context identifies OpenClaw or OPENCLAW_CLI=1 / OPENCLAW_SHELL. → openclaw.md
    • Grok Build — context identifies Grok Build / xAI's CLI. → other.md, using the vpai CLI workflow.
  2. Only if step 1 did not match — no CLI host identity, no shell:
    • Claude Cowork — Cowork workspace, .plugin install, connector store. → cowork.md
    • Claude Chat — claude.ai or Claude desktop MCP connector (not Cowork). → claude-chat.md
  3. Fallback: terminal, scripts, CI, or unknown host → other.md

Never infer platform from MCP tool names alone (mcp__claude_ai_*, mcp__*__fetch-entities, etc.). Tool prefixes are not a host signal when step 1 already identified Claude Code, Codex, OpenClaw, or Grok Build. A generic CLI Environment block (cwd, shell, platform) establishes shell access, not host identity; without an explicit host match, use the fallback guide.

PlatformRead now
Claude Codeclaude-code.md — vpai CLI
OpenAI Codexcodex.md
OpenClaw gatewayopenclaw.md
Grok Buildother.md — vpai CLI
Claude Coworkcowork.md — MCP connector; not Claude Code
Claude Chat (claude.ai, Claude desktop)claude-chat.md — MCP connector; not Cowork or Claude Code
Other (terminal, scripts, CI, generic hosts)other.md

CLI hosts use vpai — each platform guide is self-contained (Sample Gate, stdout/csv_path, CSV upload). Chat/Cowork follow each MCP tool's description and input schema only.

Hard Rules

  1. tool_reasoning (or platform equivalent) on every real call. Use the user's request verbatim. Reuse across the whole workflow. Skip only when inspecting a tool's input schema with no real execution.
  2. Chain via session + CSV, never paste IDs. Each step returns session_id and csv_path (names may vary slightly by host). Never invent session_id — on the first step omit it and use the value MCP returns; on later steps copy it exactly from the prior tool JSON. Pass --csv-path (CLI) from the prior csv_path. Do not paste raw ID lists into tool arguments when the platform reads IDs from the source CSV. For fetch-entities prospects scoped to earlier companies, pass the prior business output's csv_path via --businesses-csv-path per your platform guide.
  3. You may edit the CSV between steps — keep the ID column. On CLI hosts, read the file at csv_path, filter/dedupe/add columns, or rewrite rows, then pass the result to the next step via --csv-path. Always retain the entity ID column: business_id for business workflows, prospect_id for prospect workflows. Every row you want processed next must have a non-empty ID in that column. Do not rename or drop those headers — the CLI reads them verbatim for ID injection and joins.
  4. autocomplete first for: naics_category, linkedin_category, company_tech_stack_tech, job_title, interests, skills, business_intent_topics, city_region. Use returned standardized values, not raw user wording.
  5. Never invent tool parameters. Before the first real execution of each distinct tool in a workflow, read that tool's live input schema (from the connector or your platform's schema discovery). Do this even when the planned call looks obvious — schemas drift. Re-read when tools, filters, or shapes change materially. Build each call only from fields confirmed by that schema.
  6. Session continuity. Reuse session_id from prior tool JSON on every subsequent step in the same workflow — never make one up. Reuse the prior step's csv_path via --csv-path (CLI) when chaining enrich, events, or scoped fetch. Start a new session only for a genuinely new task.
  7. fetch-entities-statistics only when stats supports the full fetch. Compare your planned fetch-entities payload to the input schema for fetch-entities-statistics. Call statistics only if every filter key, value shape, entity_type, and scope you rely on (including session-scoped business CSVs) is accepted the same way as for fetch-entities. If any part is missing, unsupported, or needs a different shape, skip stats — do not call it with a partial or guessed subset. When you do call it, reuse the same filter object (and supported scope) as fetch-entities, plus tool_reasoning where required. Call stats again before a full-scale fetch if filters or scope changed and the full fetch filter set still fits statistics.

Tool Mechanics

Behavior and caveats the live input schema may not spell out. For allowed filter values, enum strings, and enrichment type names, always use each tool's input schema.

Autocomplete

Requires autocomplete: linkedin_category, naics_category, company_tech_stack_tech, job_title, interests, skills, business_intent_topics, city_region.

Does not require autocomplete: company_country_code (ISO Alpha-2), company_region_country_code (ISO 3166-2), fixed buckets (company_size, company_revenue, company_age, job_level, job_department — exact allowed strings come from fetch-entities / fetch-entities-statistics input schema), website_keywords (free text).

Mutual exclusions: linkedin_category and naics_category — use one, not both. company_region_country_code and company_country_code — use one.

Picking values: Autocomplete may return noisy variants. Pick the canonical clean value (usually the first clean result). Multiple values broaden with OR logic; avoid near-duplicates unless you want a wider search.

Fetch

  • job_title is substring-match, not exact. For executive searches, combine with job_level (usually c-suite) to remove assistants, advisors, and office-of roles.
  • company_size uses fixed buckets with no exact numeric cutoff. For "over N employees", approximate with adjacent buckets or enrich with firmographics for exact headcount.
  • Business location filters (company_country_code, company_region_country_code, city_region) match headquarters only — not branch or operating locations.
  • Prospects at prior companies: For fetch-entities with entity_type: prospects scoped to businesses from an earlier step, chain via session and pass the prior business step's csv_path as --businesses-csv-path per your platform guide.
  • Row limits and pagination: per your platform guide (CLI: --number-of-results; Chat/Cowork: MCP tool docs).
  • max_per_company: auto-applied on prospect fetches so results spread across companies (not all from one big employer); omit from --args unless the user sets it — their value overrides the auto cap.

Events (fetch-businesses-events / fetch-prospects-events)

  • Chain from the prior step's session and source CSV with business_id (business events) or prospect_id (prospect events).
  • When IDs come from the source CSV, do not also pass business_ids / prospect_ids inline.
  • Batching is capped at 20 IDs per request; hosts may chunk and merge automatically.
  • Output columns are event_<event_type>: each cell is a JSON array string for that type (newest-first, capped per type), or empty when no events exist for that type.

Match

  • Skip match if you already ran fetch-entities — those results include IDs.
  • CSV file upload is CLI only — see your platform guide.

Enrich

  • financial-metrics requires parameters.date (see input schema for format).
  • website-keywords requires parameters.keywords.
  • Enrichment alone does not find people at a company — use fetch-entities with entity_type: prospects (scoped to prior companies when needed) instead.
  • When chaining from a source CSV, business_ids / prospect_ids need not appear in tool arguments.

Filter Pattern

{ "values": ["v1", "v2"], "negate": false }   // include or exclude
{ "gte": 6, "lte": 24 }                       // range
true | false | null                           // boolean (not wrapped)

Exception — business_intent_topics: use { "topics": ["Category:Topic"], "negate": false } (not values). Topics must come from autocomplete.

Location matching: Business location filters (company_country_code, company_region_country_code, city_region) match a company's headquarters only — not branch/operating locations. A search for "companies in the UK" returns companies HQ'd in the UK, and excludes e.g. a foreign company that merely operates there. This is the default for all fetch-entities / fetch-entities-statistics business queries and is not user-configurable.

Limits

ToolLimit
match-business50 per call
match-prospects40 per call
enrich-business50 IDs per call
enrich-prospects50 IDs per call
fetch-businesses-events / fetch-prospects-eventsUp to 20 IDs per request when batching. Pass event_types and timestamp_from. When chaining, IDs come from the source CSV — not inline ID arrays.

Troubleshooting

ErrorSolution
Auth, connector setup, CLI invocation, flags, or chaining syntaxRead the matching platform guide from the Platform table
Empty resultsCheck filter values; run autocomplete for controlled-vocab fields; re-read the tool's input schema
linkedin_category + naics_category togetherMutually exclusive — use one
Invented or unconfirmed parametersRe-read the live input schema before calling; build arguments only from confirmed fields

Support

When the user asks a support-related question (for example: how to contact support, who to contact, they have an issue, or something isn't working), reply briefly and point them here — do not dig into Explorium internals or over-explain.

For support, please reach out to the Vibe Prospecting team via the contact page:

👉 https://www.vibeprospecting.ai/contact-us

Commands 1

Skip sample (this request only)

Hosts: Claude Code, Codex, OpenClaw, and other CLI hosts only (claude-code.md, codex.md, openclaw.md, other.md). Not Claude Chat or Cowork — those have no Sample Gate.

The user's query is:

$ARGUMENTS

First: load the normal Vibe Prospecting skill (vibe-prospecting / /vpai:vibe-prospecting), detect the host, and read the matching CLI platform guide. Follow that skill for install, auth, schema discovery, autocomplete, chaining, and export rules.

Then, for this user turn only, override Sample Gate:

  1. Do not run the 5-entity sample preview.
  2. Do not wait for approval before full scale.
  3. Run the complete prospecting workflow at the user's requested scale immediately.
  4. This does not persist — later messages without an explicit skip return to sample-first mode.
  5. After full scale, copy csv_path to the working directory with a proper name based on the user's question.

Treat $ARGUMENTS as the normal prospecting request. If $ARGUMENTS is empty, ask what they want to fetch or enrich at full scale.

Sponsors

Details

Marketplace
explorium-ai/vibeprospecting-plugin
Source
explorium-ai/vibeprospecting-plugin
Category
Data
License
MIT
Keywords
b2b, prospecting, leads, sales, enrichment, companies, contacts