What is ctxd?

March 17, 2026

ctxd stands for context daemon. A daemon is a background process that runs continuously, handling requests without human intervention. That is exactly what ctxd does for your files.

You connect your cloud apps. ctxd runs in the background. It ingests, parses, indexes, and syncs your data. When an AI agent needs context, it queries ctxd. No human in the loop.

The problem ctxd solves

Your files live in Google Drive, Slack, GitHub, and Calendar. Your AI agents cannot see them.

LLMs are powerful reasoners, but they are blind. They have no access to your spreadsheets, your Slack threads, your GitHub PRs, or your meeting notes. Your most important data sits behind OAuth walls in formats LLMs cannot consume.

→

ctxd_

Parse. Index. Search. Sync.

→

Any AI agent

API, CLI, or MCP

Every AI application that wants to reason over user data has to solve the same problems from scratch. Connect to APIs. Handle OAuth. Download files. Parse PDFs. Build search. Keep everything in sync. That is months of infrastructure before any reasoning happens.

So most AI tools punt. They ask you to copy-paste into a chat window. Or they skip private data entirely.

What ctxd is

ctxd is cloud context infrastructure. The layer between your files and your AI agents.

Connect your apps and ctxd handles the rest. Files are downloaded, parsed (including OCR for PDFs and images), and indexed. The result is clean, searchable text with metadata. Any AI agent can query it.

ctxd is not an assistant. It does not answer questions. It does not generate text. It provides the context that makes assistants useful.

How the name works

ctx = context. d = daemon.

In Unix, a daemon is a long-running background service. httpd serves web pages. sshd handles secure connections. ctxd serves context.

The analogy is deliberate. ctxd runs continuously. It watches your connected apps for changes. New file uploaded to Drive? Indexed. Message posted in Slack? Indexed. PR merged on GitHub? Indexed. No cron jobs. No manual triggers.

Four ways to use it

ctxd exposes context through four interfaces. All four hit the same index.

CLI / for developers

Search from the terminal. Pipe results into scripts. Connect apps without leaving your shell.

API / for your code

REST endpoints for search and retrieval. Build context into any application.

from ctxd import Client

client = Client()
results = client.search(
    "Q4 revenue projections",
    apps=["google-drive", "slack"]
)

for r in results:
    print(f"{r.title} ({r.app})")

MCP / for agents

Any MCP-compatible agent calls ctxd natively. The agent searches your files as part of its reasoning loop.

Your agent

ctxd

MCP server

Your apps

Skills / for automation

Composable actions your agents can run. Defined in a SKILLS.md file. Extend what ctxd can do beyond search.

What makes ctxd different

Built for agents, not dashboards. Most search products have a UI. Your agents cannot call a UI. ctxd exposes search via API, CLI, and MCP. Anything that writes code can use it.

Programmatic search. ctxd has a typed DSL. Agents filter by date, app, channel, or folder. They combine filters with boolean logic. Queries compile to an AST, not string matching.

All connectors included. No per-connector fees. Google Drive, Slack, GitHub, and Calendar are included at every pricing tier.

Minutes to deploy. OAuth handles authentication. Indexing runs automatically. No ETL pipeline. No schema to define.

Permission-aware. Every query is scoped to the authenticated user. No cross-tenant leakage. Every result links to its source.

Who it is for

AI application builders. Teams building agents, copilots, or assistants that need user data. Plug into ctxd instead of building connectors from scratch.

Teams using AI tools. If your team uses Claude, ChatGPT, or Cursor, ctxd gives those tools access to your internal knowledge.

Knowledge-heavy teams. Engineering, product, legal, research. Any team whose work depends on finding information across Drive, Slack, GitHub, and Calendar.

The short version

ctxd is a daemon that turns your cloud files into searchable context for AI agents. Connect your apps. Index everything. Query from anywhere.