bdr-prospect

Estrellas12

Forks6

Actualizado21 de mayo de 2026 a las 05:04

Research target accounts, enrich leads, draft personalized cold-email openers, and queue Gmail drafts for review

Instalación

Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.

Fuente

shogo-labs

shogo-labs/shogo-ai

Abrir repositorio de GitHub Ver repositorios del creador

Descarga

Ocupaciones relacionadasSOC

Basado en la clasificación ocupacional SOC

TelevendedoresVentas y ocupaciones afines·SOC 41-9041

SKILL.md

readonly

name	bdr-prospect
version	1.0.0
description	Research target accounts, enrich leads, draft personalized cold-email openers, and queue Gmail drafts for review
trigger	bdr\|prospect\|cold email\|cold outreach\|outbound\|lead list\|enrich\|series a\|founders
tools	["tool_search","tool_install","web","browser","memory_read","memory_write","send_message"]

BDR Prospecting

When triggered, build or extend the BDR Pipeline:

Confirm ICP — Read MEMORY.md for saved ICP. If the user supplies new criteria, persist it. Required fields: industry, stage, geography, role, recency window.
Source leads — Use web and browser to find candidates that match the ICP. Validate stage and geography from a primary source (Crunchbase, official press release, SEC filing, company blog) — never just a LinkedIn headline.
Build canvas — Use write_file to create a BDR Pipeline canvas with a Lead table:
- Fields: name, role, company, companySize, stage, fundingDate, location, email, linkedin, recentSignal, signalSource, opener, draftStatus (none|drafting|queued|sent|replied|bounced), gmailDraftId, notes
Enrich rows — Add leads in batches of 10 by rewriting the canvas file. For each row, capture a real, dated recentSignal plus the URL it came from. If you cannot find a real signal, set draftStatus to none and notes to "needs research" rather than inventing one.
Draft openers — For each enriched row, write a personalized opener (under 90 words) that opens with the signal, ties it to the user's value prop, and ends with a low-friction CTA. Update the row's opener field.
Connect Gmail — Check via tool_search({ query: "gmail" }). If missing, tool_install({ name: "gmail" }) so the user can OAuth. Do not proceed to drafts until Gmail is connected.
Queue Gmail drafts — Once Gmail is connected, call GMAIL_CREATE_DRAFT for each row using the user's sender identity. Save the returned draft id to gmailDraftId and set draftStatus to queued. Never send.
Persist — memory_write a summary: ICP used, total rows added, drafts queued, anything skipped and why.
Notify — send_message a one-line summary so the operator knows the batch is ready for review.

If Gmail isn't connected, build the pipeline and openers anyway, but stop before draft creation and tell the user exactly which step is blocked and how to unblock it.

Más de este repositorio

mismo repositorio

business-intake

shogo-labs/shogo-ai

Collect a local business's REAL details in one pass, then fill the site — so nothing is ever invented

2026-07-0412

store-intake

shogo-labs/shogo-ai

Collect a shop's REAL products and details in one pass, then fill the catalog — so nothing is ever invented

2026-07-0412

stripe-checkout

shogo-labs/shogo-ai

Switch on the store's real Stripe payments — the hosted-checkout route already ships; this wires the owner's key and verifies it

2026-07-0412

dev-activity-track

shogo-labs/shogo-ai

Fetch GitHub developer activity (commits, PRs, reviews) and build an activity dashboard

2026-05-2112

github-ops

shogo-labs/shogo-ai

Monitor GitHub repos — fetch PRs, issues, CI status via Composio and display on canvas

2026-05-2112

habit-track

shogo-labs/shogo-ai

Track daily habits on a canvas kanban board with streaks

2026-05-2112

BDR Prospecting

When triggered, build or extend the BDR Pipeline:

Confirm ICP — Read MEMORY.md for saved ICP. If the user supplies new criteria, persist it. Required fields: industry, stage, geography, role, recency window.

Source leads — Use web and browser to find candidates that match the ICP. Validate stage and geography from a primary source (Crunchbase, official press release, SEC filing, company blog) — never just a LinkedIn headline.

Build canvas — Use write_file to create a BDR Pipeline canvas with a Lead table:

Fields: name, role, company, companySize, stage, fundingDate, location, email, linkedin, recentSignal, signalSource, opener, draftStatus (none|drafting|queued|sent|replied|bounced), gmailDraftId, notes

Enrich rows — Add leads in batches of 10 by rewriting the canvas file. For each row, capture a real, dated recentSignal plus the URL it came from. If you cannot find a real signal, set draftStatus to none and notes to "needs research" rather than inventing one.

Draft openers — For each enriched row, write a personalized opener (under 90 words) that opens with the signal, ties it to the user's value prop, and ends with a low-friction CTA. Update the row's opener field.

Connect Gmail — Check via tool_search({ query: "gmail" }). If missing, tool_install({ name: "gmail" }) so the user can OAuth. Do not proceed to drafts until Gmail is connected.

Queue Gmail drafts — Once Gmail is connected, call GMAIL_CREATE_DRAFT for each row using the user's sender identity. Save the returned draft id to gmailDraftId and set draftStatus to queued. Never send.

Persist — memory_write a summary: ICP used, total rows added, drafts queued, anything skipped and why.

Notify — send_message a one-line summary so the operator knows the batch is ready for review.

If Gmail isn't connected, build the pipeline and openers anyway, but stop before draft creation and tell the user exactly which step is blocked and how to unblock it.