iGPT: The context engine for work

What

iGPT is an email intelligence API designed to give AI agents and automations usable context from full email threads, files, and attachments through a single request. The product is positioned for developers and enterprise teams that want structured, automation-ready outputs without building a separate retrieval pipeline, prompt chain, or parsing stack.

Based on the page, the core workflow is: connect email sources, let iGPT index messages and attachments, then send a natural-language request to one endpoint that handles retrieval, context shaping, and reasoning. It appears aimed at teams building email assistants, workflow automation, support copilots, CRM agents, and other systems that depend on current, traceable email context.

Features

• Unified email context API — A single endpoint combines retrieval, context shaping, and reasoning so developers can query email data without assembling multiple services.
• Deep indexing of emails and attachments — The platform automatically indexes messages, files, PDFs, and spreadsheets to make buried context easier to access.
• Thread reconstruction — It rebuilds email conversations across time, participants, and attachments, which helps preserve decision history and conversational continuity.
• Hybrid retrieval with reranking — Semantic search, keyword search, and filters are combined and scored in one call to improve relevance across messy email datasets.
• Schema-based structured outputs — Developers can request results in a defined JSON schema, which is useful for turning email context into tasks, approvals, action items, or other machine-readable outputs.
• Cited answers and auditability — Responses link back to source messages and attachments, supporting traceability for review, verification, and downstream workflows.

Helpful Tips

• Validate source coverage early — Before adopting a product in this category, confirm which email systems and attachment types are actually supported, since this page does not list all connector details.
• Start with high-value extraction use cases — Action items, approvals, ownership, deadlines, and support history are practical initial workflows because they fit structured outputs well.
• Test citation quality, not just answer quality — For enterprise use, review whether cited messages and attachments are sufficient for users to verify decisions and reasoning.
• Check deployment and data-handling fit — This page mentions cloud, hybrid, and private deployment options plus zero data training and zero retention claims, so buyers should map those options to internal security and architecture requirements.
• Measure workflow reliability on real thread complexity — Products in this space are most valuable when they can handle long threads, forwarded chains, mixed participants, and attachment-heavy conversations consistently.

OpenClaw Skills

Within the OpenClaw ecosystem, iGPT could likely serve as a context layer for agents that need reliable access to email-based business memory. Likely skill patterns include an email action-item extractor, a board-meeting follow-up agent, a support thread summarizer, a contract-and-approval tracker, or a CRM timeline builder that turns thread history into structured records with citations.

This combination could be especially useful in operations, support, sales, legal, and executive workflow environments where decisions are scattered across inboxes and attachments. If OpenClaw agents orchestrate multi-step workflows, iGPT could likely provide the evidence-backed context step: retrieve the relevant thread, extract structured facts, and hand them to downstream skills for task creation, escalation, reporting, or human review. The page does not confirm a native OpenClaw integration, so this is best understood as a likely orchestration use case rather than a stated product capability.