Overview

Use this skill as Kelly's inbound-inquiry reception desk for foreign-trade sales. Inquiries arrive via WhatsApp, Instagram, Messenger, and email; sales reply late and leads leak. The AirApp aggregates them into one Busabase pipeline (new → replied → quoted → negotiating → won/lost), maintains a product knowledge base + FAQ so the agent can draft accurate replies and quotes, enforces approval before anything is sent, and surfaces stale deals past the follow-up SLA. Collecting real inquiries and sending real replies/quotes are both genuine external-platform operations a browser cannot perform (no secrets, no outbound platform calls): scripts/ingest_inquiries.mjs is the single write-path for agent-collected or manual inquiry payloads, scripts/sync_products.mjs imports/refreshes the product knowledge base, and scripts/send_approved.mjs sends approved replies/quotes. The AirApp itself only reads Busabase and writes queued drafts/decisions/follow-ups/quote edits.

Default behavior is AirApp-first. Unless the user explicitly asks only for explanation, ingest whatever inquiries/products are collected and give the user the clickable AirApp URL (or the local preview URL when local preview is explicitly requested). Use chat-only mode only when the user says "纯聊天", "chat only", "不要打开 UI", or similar.

The AirApp itself never talks to WhatsApp/Instagram/Messenger/email. It reads and writes Busabase records only. Both external-platform directions are genuinely trusted-process-only: scripts/ingest_inquiries.mjs is the only place inquiry payloads are written, and scripts/send_approved.mjs is the only place that sends a real reply/quote — always after a human approval recorded in Busabase.

Mandatory Dependencies

Read and follow $kelly-app-skill-creator for product behavior, visual quality, responsive layout, and the complete canonical content/kelly-inquiry-app/ artifact.
Read and follow $busabase for connection, target Space, node discovery, ChangeRequests, review, and merge behavior.
Read and follow $busabase-app-creator for resource modeling, AirApp runtime limits, security, validation, and deployment.

If a dependency is unavailable, preserve this skill's local artifact and product contracts, stop before the unavailable Busabase operation, and report the exact missing dependency. Do not invent a second data backend.

Boundary

The AirApp reads and writes Busabase records only. It must never call WhatsApp/Instagram/Messenger/email APIs, or perform any other external side effect. It cannot send anything: the composer queues drafts into the approvals Base, the approvals view records verdicts.
Every outbound reply AND quote is approval-required. Only scripts/send_approved.mjs --send sends, and only items whose status is approved; the dry run (no --send) only prints a plan.
Own accounts only: read and send exclusively through accounts the user owns and has configured. Respect each platform's terms of service and rate limits; prefer official APIs; keep collection read-only against the platform (Busabase is the only thing the ingest/sync scripts write to).
Never store passwords, QR-login payloads, or session tokens — anywhere, including Busabase. Accounts store only the channel, connector, and the names of env vars holding tokens, never the token values. For browser_agent collection the agent drives the user's own already-authenticated web session and stores only inquiry text needed for review.
Product and pricing data stays in Busabase. Treat customer contacts, conversation excerpts, price floors, and quotes as sensitive: never commit real tokens, customer exports, or Busabase credentials.

Busabase Resources

Eight Bases under one application Folder (kelly-inquiry), declared in content/kelly-inquiry-app/app/js/config.js and the generated template sidecars under content/:

accounts: connected channels — WhatsApp/Instagram/Messenger/email, connector, and env-var names for tokens (never values), status, last sync.
inquiries: the sales pipeline — customer (name/company/country/source), product interest, linked product-ids/quote-ids, stage, value estimate, owner, next-follow-up, the send target (provider-conversation-id), and an optional agent-suggested reply.
messages: one row per conversation message, joined onto its inquiry by inquiry-id.
products: the product knowledge base — SKU, MOQ, price-min/price-max, lead time, specs and faq (JSON-encoded).
quotes: quote worksheets — line items (JSON-encoded), currency, validity, terms, pricing notes, and the min-price-guard pricing-alerts (recomputed live, never trusted from storage).
approvals: the review queue — outgoing reply/quote drafts, workflow status, the human verdict fields (decision-action/decision-comment/decided-at), and the execution result (execution-status/execution-operation/execution-connector/execution-target/execution-detail/executed-at) written by scripts/send_approved.mjs.
sync-log: append-only history of ingest/sync runs per account.
settings: one row (record-id: "config") with quote defaults (currency, validity days, incoterm, payment terms, min-price guard), follow-up SLA days per stage, reply style, and the product KB source path.

Resources provision lazily through an idempotent Busabase ChangeRequest the first time the app runs in a Space; see references/inquiry-schema.md for exact field shapes. Per-inquiry rollups (last_message_at, last_incoming_at, the new→replied stage heuristic), quote totals, the min-price guard, follow-up staleness, and metrics are all recomputed client-side from inquiries/messages/products/quotes on every read — never stored.

First Run And Onboarding

On invocation, check the accounts Base. If empty, guide setup before collecting real inquiries.

Onboarding asks, turn by turn:

Which channels receive inquiries (WhatsApp / Instagram / Messenger / email) and which connector method per account (see Collection Workflow). Ask for non-secret details only: channel, display name, handle, and which env var names hold the tokens. Never ask the user to paste secret values into chat; secrets belong only in local env files.
Product KB import: a JSON or CSV file of products (SKU, MOQ, price range incl. price_min floors, lead time, specs, FAQ), imported via scripts/sync_products.mjs.
Quote defaults: currency, validity days, incoterm/payment terms, and whether the min-price guard is enabled.
Follow-up SLA days per stage (defaults: new 1, replied 2, quoted 3, negotiating 5).
Reply style: tone, language policy, signature, and "do not say" guardrails.

Write the answers to the settings Base's single record-id: "config" row (quote-defaults/follow-up/reply-style/kb-source-path, each JSON-encoded where structured) via busabase-sdk. Register an account with:

node skills/kelly-inquiry/scripts/ingest_inquiries.mjs onboarding-payload.json --apply

where onboarding-payload.json carries an account object (see references/inquiry-schema.md; inquiries can be omitted on this first run).

Local App

Default behavior is AirApp-first — give the user the clickable AirApp URL. Start pnpm --dir content/kelly-inquiry-app dev only when local preview/debugging is explicitly requested.

Required app views (hash routes):

#/overview: inquiry command desk. Human-attention panel (replies/quotes awaiting approval, unanswered new inquiries, stale deals past the follow-up SLA), KPI cards (inquiries this week with channel badges, reply median time, quotes sent, win rate), a pipeline funnel summary (inline SVG bars: new→replied→quoted→negotiating→won), and an oldest-unanswered indicator.
#/inquiries and #/inquiries/<id>: the pipeline table — customer, country flag/code, channel badge, product interest, stage, value estimate, last message age, next follow-up date, owner. Detail: conversation excerpt (bubbles), customer profile (company, country, source), linked products, quote history, an agent-drafted reply with editable text plus Queue reply (goes to Approvals), and a follow-up scheduling field.
#/quotes and #/quotes/<id>: quote worksheet — quote no, customer, line items, currency, validity, status (draft/sent/accepted/expired/declined). Detail: editable line items sourced from the product KB (draft quotes only), terms, and agent pricing notes with the min-price guard result from the KB floors.
#/approvals: the review queue with workflow states needs_review / changes_requested / approved / done / blocked over outgoing replies AND quotes. Each item shows target channel + customer, the draft, reason/context, editable text, decision buttons (approve / request changes / save edit / block), and stable refs (Reply #1 / Quote #2). done means sent, with the execution result written by scripts/send_approved.mjs.
#/products and #/products/<id>: the product KB — cards with name, SKU, MOQ, price range, lead time, FAQ count; detail with specs and the FAQ entries the agent uses for drafting.
#/settings: sanitized config — channels/accounts with connector method + env readiness booleans, product KB source, quote defaults (currency, validity days, min-price guard), follow-up SLA, sync log, and onboarding state. Never secrets.

Demo mode:

?demo=overview, ?demo=inquiries, ?demo=quotes, ?demo=approvals, ?demo=products, and ?demo=detail (opens the featured hot WhatsApp inquiry wa-mueller-led-panels with a drafted reply and a draft quote) select named deterministic mock scenes. Persona: "Lumina Lighting Co.", a foreign-trade LED-lighting supplier.
lang=en or lang=zh forces UI chrome language for screenshots. With lang=zh the chrome AND agent-generated content (reasons, notes, product names, pricing notes) are localized; drafted replies and conversation quotes stay in the buyer's language. Deep links such as /?demo=detail&lang=zh#/inquiries/wa-mueller-led-panels work.
Demo mode never reads or writes Busabase. Composer, approvals, follow-up, and quote-edit buttons still work but act on in-memory state only (running the exact same refreshInquiryDerived/recomputeQuoteTotals/applyMinPriceGuard functions the Busabase provider uses) and show a demo notice.

UI language: English and Chinese chrome with Auto default following the browser language; explicit selector persisted locally. Keep customer names, message content, and drafted outbound text in their original language.

File Contract

Read references/inquiry-schema.md before editing the app, scripts, or Busabase field shapes.

Collection Workflow

Detect mode. Default to AirApp-first.
Check the accounts Base. If empty, enter onboarding.
Collection reuses the connector reality documented in kelly-messenger. Per account, declare a connector:
- whatsapp_cloud — WhatsApp Business Cloud API (access_token_env + phone_number_id_env). Inbound messages arrive via webhook only, so history is collected via ingest payloads; sends use the Cloud API.
- instagram_graph / messenger_graph — Meta Graph API for professional-account DMs and Page messages (access_token_env plus ig_user_id_env / page_id_env).
- email_agent — hand off to the kelly-email skill: it collects inquiry emails, the agent normalizes them into an ingest payload; sends go back through kelly-email drafts.
- browser_agent — the agent drives the user's own already-authenticated web session (e.g. WhatsApp Web, Instagram web) with the browser skill, then writes a payload. No passwords or QR secrets are ever stored.
- manual — the user or agent prepares an ingest payload by hand.
All collected data enters through one write path: node scripts/ingest_inquiries.mjs payload.json --apply. It validates the payload, dedupes by stable inquiry/message ids, merges into Busabase, applies the stage heuristic (an outgoing reply promotes new → replied), upserts the account, and appends a sync-log entry. Omit --apply first to see a dry-run summary.
While drafting, the agent may attach a suggested-reply per inquiry (prefilled in the composer) and queue reply/quote drafts into the approvals Base with suggested-by: "agent" and a clear reason — always grounded in the product KB and reply style, never below price_min.
Give the user the AirApp URL (or local preview URL). Surface connector problems as printed warnings, not silent failures.

Quoting Workflow

Ground every quote in the product KB: SKU, MOQ, tier pricing inside price-min–price-max, lead time, and FAQ facts (certificates, OEM options, dimming, packaging). Import/refresh the KB with node scripts/sync_products.mjs products.json|products.csv --apply (zero-dependency CSV parser with quoted-field support).
Min-price guard: config quote-defaults.min_price_guard plus per-product price-min floors. Any line priced below its floor raises a pricing_alerts entry (recomputed live by applyMinPriceGuard); with block_below_price_min the agent must not queue such a quote for sending — block it and ask the user instead.
Build the quote as a draft in the quotes Base (quote no, line items, currency, validity from validity_days, terms from quote defaults, pricing notes explaining the tier used and the guard result) and queue a matching kind: "quote" approval item referencing it via busabase-sdk.
The user edits draft line items in #/quotes/<id> (the provider recomputes totals via recomputeQuoteTotals and re-runs the guard) and gives the verdict in #/approvals.
After a quote is sent, scripts/send_approved.mjs sets its status to sent; track accepted / expired / declined from the conversation, and move the inquiry stage accordingly (quoted, negotiating, won, lost).

Approval And Send Workflow

scripts/send_approved.mjs is the executor — there is no separate execute_decisions script.

Queue: the user writes or edits a reply in the composer (optionally starting from the agent's suggested-reply) and clicks Queue reply; the app writes it to the approvals Base as needs_review via busabase-sdk. The agent queues its own reply/quote drafts the same way.
Review: in #/approvals the user approves, edits (Save edit), requests changes, or blocks each item — written directly onto the approval record. From a standalone local preview the write merges immediately (trusted operator); from the deployed AirApp it creates a pending ChangeRequest for the trusted process to merge.
Agent revision loop: for an item moved to changes_requested, redraft the text honoring the decision-comment, the config reply_style, and the KB (re-check the min-price guard for quotes), then write it back to needs_review.
Send: only after the user asks to send, run node scripts/send_approved.mjs (dry-run) and show the plan — planned sends, targets, and missing-token blockers. With explicit approval, run node scripts/send_approved.mjs --send: it re-reads Busabase immediately before sending, sends API-connector items via the official APIs (WhatsApp Cloud, Instagram/Messenger Graph), marks email_agent/browser_agent/manual items as handoff_to_agent for the agent to deliver (kelly-email drafts, or the user's own web session), sets sent items to done, and writes the execution result onto each approval record.
Report per-item results back to the user with the stable Reply #N / Quote #N refs.

Follow-Up Reminders

Stale deals are how leads leak; the agent owns catching them.

On every invocation (and after every ingest), compare each active inquiry (new/replied/quoted/negotiating) against follow-up.sla_days for its stage and its next-follow-up date. This is computed live via staleInquiries/isFollowUpOverdue — never trusted from a stored flag.
For each overdue deal, draft a follow-up reply into the approvals Base (suggested-by: "agent", reason citing the SLA breach and deal value) so the human can approve before anything is sent.
The UI surfaces the same signal: the sidebar stale-deal counter, the overview "Stale deals past follow-up SLA" panel, and red overdue dates in the pipeline table. The user can reschedule with the follow-up field in the inquiry detail.
When the user asks "what should I chase today", answer from the stale list (staleInquiries) plus any queued follow-up drafts, with Reply #N refs for anything already drafted.

Safety Defaults

Never send without an approved status recorded in the approvals Base, and never bypass the dry-run → --send sequence.
Never quote below a product's price-min without an explicit human decision recorded on the item.
Prefer read-scoped tokens where the platform offers them; keep collection strictly read-only against the platform.
Redact tokens and token-like strings from logs, reports, and UI state; expose only env-var readiness booleans.
Keep sends idempotent: stable item ids, execution results stored on the approval record, and re-reading Busabase before each send.
If a send target is missing (provider-conversation-id), leave the item approved with an execution-status: error and ask for configuration instead of guessing.
Honor platform rate limits; on 429s back off rather than retrying aggressively.

Useful Commands

node skills/kelly-inquiry/scripts/ingest_inquiries.mjs payload.json --apply
node skills/kelly-inquiry/scripts/sync_products.mjs products.csv --apply
node skills/kelly-inquiry/scripts/send_approved.mjs
node skills/kelly-inquiry/scripts/send_approved.mjs --send
pnpm --dir skills/kelly-inquiry/content/kelly-inquiry-app dev

In normal use, invoke /kelly-inquiry, let the skill ingest the configured accounts/products, and open the AirApp.

kelly-inquiry

Kelly Inquiry

App UI Screenshots