OpenCues enables native AI integration anywhere you type. Model agnostic and fully open source. Inline agents and prompting.

Rather than navigating to a chat interface or AI-enabled input box, define a query in plain text: an LLM discovers your query and answers it inline, right where you wrote it.

OpenCues is platform, model, and provider agnostic, engineered from the ground up to enable native inline AI.

You type	You get
hey can u send me that report when u get a sec make this formal _	Could you please send me that report at your earliest convenience?
4 + 4 = _	4 + 4 = 8
hello world translate to japanese _	こんにちは世界
draft an email to my landlord asking for a rent reduction _	(the email, written)
ffmpeg command to convert a video to web-ready mp4 _	ffmpeg -i input.mov -c:v libx264 -preset slow -crf 23 -c:a aac -b:a 128k -movflags +faststart output.mp4

Quickstart

Name: opencues
Author: opencues

⚠️ Private beta. npm install -g opencues isn't live yet. Install from a clone instead — see BETA-INSTALL.md. Everything past the install step below is accurate today.

npm install -g opencues              # see BETA-INSTALL.md until this is live
opencues set-key cerebras csk-...    # cerebras.ai — free tier, lowest latency
opencues install claude-code         # or: opencode | gemini-cli | chrome | shell
claude-cues                          # launch — native `claude` is untouched

Full walkthrough, prerequisites, and per-host detail: docs/install.md. opencues doctor diagnoses anything that looks wrong.

Integrations

Host	Status	Install
Claude Code	Available	`opencues install claude-code`
OpenCode	Available	`opencues install opencode`
Gemini CLI	Beta	`opencues install gemini-cli`
Chrome	Beta	`opencues install chrome`
Shell	Beta	`opencues install shell`

Each pins its own upstream fork and never touches your native host install.

⚠️ Windows: not supported natively, run inside WSL2 (see docs/install.md).

What you get

Feature	What it does
Blanks	Type `_` for free-form generation, translation, formatting, full rewrites, or keyword-bound system actions (`volume _`, `weather _`).
Sentence rewrites	Cycle a whole sentence to a different register (formal, concise, ...) seamlessly, no `_` needed.
Word cues	Navigate to a single word and cycle a smaller LLM-suggested alternative.
Personal + ambient context (opt-in)	`my email _` substitutes your real address; fluid lookups can read the page you're on.
Hot-reload	Every `.md` config picks up edits in ~2s, no restart.

Full feature catalogue (44 concepts): docs/features/README.md.

Configuration & LLM providers

Config lives at ~/.cues/ — one OPENCUES.md for runtime settings, plus per-surface source folders. Seven providers supported; set an env key or opencues set-key and you're done.

Full reference: docs/configuration.md · docs/guides/llm-providers.md (switching provider/model, free mode, failover).

Providers:

Security

OpenCues has no tool handlers or exec layer for LLM output — no MCP-tool execution, no agentic actions, no side-effect channel. Worst-case, an LLM response lands as user-visible text in the buffer you review before submitting. That single invariant is what keeps prompt injection a UX failure instead of a data-exfiltration channel, across every surface below.

23 of 27 tracked attack classes closed, 3 closed-with-caveat, 1 tracked for the future pack registry (full audit table).

Defense	What it covers
Sandbox isolation	Third-party blank JS runs in a real V8 isolate (`isolated-vm`) — its own realm, own intrinsics, no sandbox-escape via constructor-chain pivots
Capability gates	A blank only gets `network`/`llm`/`storage`/`secrets` access if declared; secrets without a matching host binding are refused at load time
Resource quotas	Sliding-window caps on fetches, LLM calls, and storage writes — no polling hammer, no runaway LLM burn
Output sanitization	Blank output is stripped of HTML/script tags, zero-width chars, and bidi overrides before it reaches the buffer

Check a pack before trusting it:

opencues review ./untrusted-pack/

Full threat model: docs/architecture/security-audit.md. Reporting a vulnerability: SECURITY.md.

Contributing

What	Where
New host integration	`docs/guides/adding-an-integration.md`
New cue/blank	`docs/guides/adding-a-cue-blank.md`
Working on the reference runtime	`CONTRIBUTING.md`
New to the terms	`docs/glossary.md`

Join the community — questions, feedback, and the people building alongside you. We're also part of OpenSourceIRL, a community for people building in the open.

Community:

License

Apache License 2.0.

Design