io.github.mgthompo1/anchorr-mcp
Agent-native CRM. 25 tools — contacts, deals, sequences, enrichment waterfall, audit log.
Que pouvez-vous faire avec io.github.mgthompo1/anchorr-mcp ?
enrich_contact— single contact via email or(first_name, last_name, domain)enrich_company— single company via domain or namebulk_enrich_contacts— up to 200 contact IDs in one call, concurrency-capped, returns aggregate stats and per-contact outcomes
Installation
Ajoutez ceci à la configuration MCP de votre client (Claude Desktop : claude_desktop_config.json ; Cursor : .cursor/mcp.json ; VS Code et les autres utilisent la même forme « mcpServers »).
claude mcp add --transport http anchorr-mcp 'https://anchorr-mcp.mgthompo.workers.dev/mcp'
{
"mcpServers": {
"anchorr-mcp": {
"type": "http",
"url": "https://anchorr-mcp.mgthompo.workers.dev/mcp"
}
}
}
Documentation
Tirée du README du projet sur GitHub, actualisée chaque semaine. Les liens ouvrent GitHub.
Sommaire
Anchorr MCP
The MCP server for Anchorr — the agent-native CRM. 25 tools across contacts, companies, deals, sequences, activities, and an enrichment waterfall (Apollo → Hunter → AI research) with shared cache and per-call audit.
Every CRM was built for humans typing into forms. We built one for AI agents executing pipeline.
Connect
| URL | https://anchorr-mcp.mgthompo.workers.dev/mcp |
| Auth | Authorization: Bearer ancr_live_… |
| API key | Mint at anchorr.app/settings → API Keys |
| Catalogue | anchorr.app/mcp — full tool docs and example transcripts |
API keys are scoped per-tool category — issue narrow keys per agent (e.g. enrichment-only for a research bot, contacts:write + sequences:write for a prospecting agent).
Tool surface
Enrichment — Resolve contacts and companies through a provider waterfall. Cache is shared with the Anchorr web product, so a hit from Cursor warms the kanban; a hit from the kanban warms the agent.
enrich_contact— single contact via email or(first_name, last_name, domain)enrich_company— single company via domain or namebulk_enrich_contacts— up to 200 contact IDs in one call, concurrency-capped, returns aggregate stats and per-contact outcomes
Contacts & Companies — list_contacts, get_contact, create_contact, bulk_create_contacts, enroll_contact_in_sequence, list_companies, get_company, create_company, update_company.
Deals & Pipeline — list_deals, get_deal, create_deal, update_deal, move_deal_stage, close_deal. Deals carry next_step and stale-deal detection.
Sequences & Templates — list_sequences, get_sequence, create_sequence, update_sequence, list_templates, create_template, update_template. Email / wait / task steps with {{first_name}} / {{company}} placeholders.
Activities — list_activities, log_activity, create_task, complete_task.
Agent — get_agent_overview, get_agent_config, update_agent_config, list_candidates, update_candidate_status. The autonomous agent that runs continuous prospecting + reply qualification on top of the CRM.
Resources
Composite reads tuned for "prep me for this meeting" / "brief me on this account" prompts. One resource fetch replaces five or six tool chains.
anchorr://org/overview— pipeline + contacts + agent + tasks snapshotanchorr://deal/{id}/brief— deal + contact + company + recent activities + researchanchorr://contact/{id}/brief— contact + company + activities + deals + sequence enrollments + researchanchorr://company/{id}/brief— company + contacts + deals + activities + research
Audit & cost
Every tool call is logged to mcp_audit_log with the calling key, parameters, success state, and duration. Every enrichment is logged to enrichment_requests with the per-attempt provider trail and cost in micro-dollars (USD × 10⁶). Cache hits are free.
Example
{
"tool": "enrich_contact",
"input": { "email": "patrick@stripe.com" }
}
{
"data": { "first_name": "Patrick", "last_name": "Collison", "title": "CEO", "company": "Stripe", "linkedin_url": "..." },
"resolved_by": "apollo",
"served_from": "provider",
"attempts": [
{ "provider": "apollo", "ok": true, "latency_ms": 412, "cost_micros": 50000 }
],
"total_cost_micros": 50000,
"total_latency_ms": 478
}
Use with Claude Desktop
{
"mcpServers": {
"anchorr": {
"url": "https://anchorr-mcp.mgthompo.workers.dev/mcp",
"headers": {
"Authorization": "Bearer ancr_live_..."
}
}
}
}
Use with Cursor
Add to ~/.cursor/mcp.json:
{
"mcpServers": {
"anchorr": {
"type": "streamable-http",
"url": "https://anchorr-mcp.mgthompo.workers.dev/mcp",
"headers": {
"Authorization": "Bearer ancr_live_..."
}
}
}
}
License
Source code in this directory is part of the Anchorr platform. The hosted MCP server is free to use with an Anchorr account; usage is metered against your Anchorr plan's quotas.
Questions fréquentes
Qu'est-ce que io.github.mgthompo1/anchorr-mcp ?
Agent-native CRM. 25 tools — contacts, deals, sequences, enrichment waterfall, audit log.
Comment installer io.github.mgthompo1/anchorr-mcp ?
Il fonctionne comme serveur distant. Dans Claude Code : claude mcp add --transport http anchorr-mcp 'https://anchorr-mcp.mgthompo.workers.dev/mcp'. Pour les autres clients, collez la configuration JSON de cette page dans leurs réglages MCP.
Faut-il une clé API pour io.github.mgthompo1/anchorr-mcp ?
Son entrée au registre n'indique aucune clé ni secret obligatoire. Certaines fonctions peuvent tout de même demander un compte auprès du service auquel il se connecte.
io.github.mgthompo1/anchorr-mcp est-il open source ?
Son code est public sur GitHub, mais le dépôt ne déclare pas de licence standard.
Qui maintient io.github.mgthompo1/anchorr-mcp ?
Il est publié par mgthompo1. AI.tn le référence sans le maintenir ni le recommander.
Issu du registre MCP officiel. Être listé n'est pas une recommandation : vérifiez le code et les permissions d'un serveur avant de lui donner accès à vos données.