Sandwalk

Integrations

Connect Slack workspaces, Modal cloud GPUs, Hugging Face, LangSmith, and scientific APIs.

Dashboard Integrations: Boltz, Fastfold AI Cloud, Modal, Slack, Vercel, Hugging Face, IBM RXN, Lambda Labs, and LangSmith

Dashboard → Integrations (/dashboard/integrations) connects external services, cloud compute providers, communication channels, and scientific APIs. Credentials are stored locally on your machine in ~/.sandwalk/config.json and automatically injected into matching tools, skills, and catalog MCP servers.


Modal
Modal
Agent Runtime · Cloud GPU

Run agent Python code and scientific pipelines on serverless CPU and GPU containers (T4, L4, A10, A100, H100) with automatic artifact sync.

modal.com ↗
Slack
Slack
Communication · Socket Mode

Connect a Slack workspace with 1-click app manifest setup so teammates can chat with Sandwalk agents in channels or DMs.

Slack Relay Agent ↗
Boltz
Boltz
Scientific API · Structure & Affinity

Biomolecular structure prediction, pocket conditioning, binder design, and binding affinity scoring.

jwohlwend/boltz ↗
FastFold AI Cloud
FastFold AI Cloud
Scientific API · Folding & MD

Managed cloud folding, BoltzGen binder design, and high-throughput molecular dynamics pipelines.

fastfold.ai ↗
Vercel (skills.sh)
Vercel (skills.sh)
Skills · Public Catalog & Audits

Search the public skills.sh leaderboard, inspect automated security audits, and install skills in one click.

skills.sh ↗
Hugging Face
Hugging Face
Data & Models · Hub Token

Authenticate with HF_TOKEN to pull gated weights, biomedical datasets, and invoke Hugging Face MCP tools.

huggingface.co ↗
IBM RXN for Chemistry
IBM RXN for Chemistry
Scientific API · Retrosynthesis

AI-driven chemical retrosynthesis, forward reaction prediction, and automated synthesis route planning.

IBM RXN ↗
Lambda Labs & RunPod
Lambda Labs & RunPod
Compute · Dedicated GPUs

Provision dedicated multi-GPU cloud instances for heavy model training, molecular dynamics, and batch screening.

Cloud GPU Compute ↗
LangSmith
LangSmith
Observability · Tracing

Optional end-to-end trace logging, latency profiling, and token inspection for LangChain and Deep Agents runs.

smith.langchain.com ↗
Lens.org
Lens.org
Literature & Patents · API

Global patent landscape search, sequence-enabled patent claims, and scholarly citation graphs.

lens.org ↗
Tavily & Exa
Tavily & Exa
Search · Web & Literature Grounding

High-precision scientific web search, paper discovery, and full-text extraction for research agents.

Web Search APIs ↗
SendGrid
SendGrid
Notifications · Email Delivery

Send automated completion alerts and HTML research digests when long-running campaigns finish.

sendgrid.com ↗

How to Configure an Integration

  1. Open Dashboard → Integrations and filter by Status (All, Configured, Not configured) or Category (Featured, API, Agent Runtime, Communication, Compute, Data, Observability, Skills).
  2. Click any integration card to open its setup view.
  3. Paste your API key or token (or use the 1-click OAuth / manifest wizard for Slack and Fastfold AI Cloud), click Validate / Save, and toggle the integration On.

Last updated on