Works with what
you already have.

ML training loops, Rake-style scripts, AI agents.

Node workers, Next.js build tasks, Deno/Bun scripts.

Backend services, devops tools, anything written in Go.

Systems services, ML infra, Rust CLIs.

Rails ActiveJobs, Sidekiq workers, Rake tasks.

Run a test suite on your phone - see passes/fails tick live.

Any Node web server - page yourself the moment a 5xx leaks.

AI assistants that speak the Model Context Protocol.

Any command that already exists - no code changes.

Cron jobs, systemd services, Makefile recipes, Docker entrypoints.

See everything Chirp is doing, live, in your terminal.

Push CI status to your phone while you grab coffee.

Self-hosted GitLab or gitlab.com - same CI story as GitHub.

Teams running CircleCI - drop-in orb, no publishing required.

Atlassian shops - same CI story via bash helpers.

Buildkite pipelines - pass/fail without touching pipeline.yml.

Containerized services - send status without adding a runtime.

Ask Cursor to send a notification - it knows how.

Same story as Cursor - Cascade learns Chirp from one file.

Long-running Claude Code sessions - phone shows Claude is working.

OpenAI Codex sessions - phone shows what tool it's running.

Copilot CLI in your repo - agent activity on the lock screen.

Cline VS Code agent - including the unique TaskResume state.

Amazon Kiro spec-driven agent - schema near-identical to Claude Code.

SST OpenCode TUI - first-class permission.asked event.

Codex CLI · Aider · Zed · any tool that reads AGENTS.md.

Daily AI API spend on your lock screen - cost figure pulled from OpenRouter's response, not a maintained rate table.

Long-running Modal functions - phone shows queued → running → done.

Replicate predictions - paste a URL on the create call, done.

RunPod serverless jobs - async completion without polling.

Inference Endpoints state changes + Spaces builds.

Together fine-tunes / batch jobs - completion without polling.

Cron miss / recovery on your lock screen - uniquely solved here.

Cronitor monitors - failure / late / recovery delivered to Chirp.

Better Stack heartbeats / monitors - incidents on your lock screen.

Languages without an SDK, embedded systems, one-offs.

Every deploy → phone. No code, no SDK.

Repos not using Actions, or global org-level events.

CircleCI users who want status without touching config.yml.

Solo founder / small team - get pinged on every sale.

Production errors - know before your users complain.

Infra alerts - CPU, error rate, latency thresholds.

Anything we haven't built a translator for - Linear, custom scripts.

Watch any service's webhooks stream live on your lock screen.