Polylane
AI agents that fix issues before they become incidents
616 followers
AI agents that fix issues before they become incidents
616 followers
Nobody should be on call. Polylane connects your code, infra and observability data, investigates every issue and opens a pull request with the fix.
This is the 7th launch from Polylane. View more
outages.live
Launched this week
Check whether a provider is down before debugging your code. outages.live brings official status pages for cloud, AI, payments, and developer tools into one feed, refreshed every few minutes. Give your coding agent outage context over MCP, browse the live feed, or subscribe via RSS, Atom, JSON Feed, or Slack. Free, with no login or API key.
Free
Launch Team
LoomHit record, hand off to AI
Promoted
Hey Product Hunt 👋
We made outages.live as a small side project at Polylane.
When an API fails or a service slows down, checking the provider's status is a useful first step before digging into your own code. We wanted to make that easy for coding agents, too.
outages.live brings official provider status pages into one feed and refreshes them every few minutes.
Add the free MCP server to Claude Code, Codex, Cursor, OpenCode, or Copilot so your agent can check for provider incidents while it works.
No login or API key needed.
You can also browse the live feed, filter by provider and impact, or subscribe through RSS, Atom, JSON Feed, or Slack.
Try it at https://outages.live.
We'd love to hear which providers we're missing and how your agent uses it. We'll be here for questions!
congrats on the launch matanya. to your question, i would use this when the agent starts blaming our code and it is actually stripe or openai being down. the providers i miss most are email ones, sendgrid and resend outages look exactly like a broken signup )
Can I set it to only notify me about services I actually use?
I'd want the MCP tool to tell the agent when an incident clears too, otherwise it either sits idle or keeps polling the feed every few minutes.
Nice! All the best on this impressive launch
This is the check I wish my agent did on its own before it starts 'fixing' things. I've watched agents rewrite working code over a provider outage more than once, so having the official status feed available over MCP — with no login or API key — feels like the right default. Curious how you handle the gap when a status page is slow to admit there's an incident, since that's usually when the agent needs it most.
giving the agent a way to check the provider before it rewrites half my code over a Stripe outage is the right instinct, that loop happens way too often
status pages are usually the last to admit anyhting is wrong tho. do you only mirror what the provider posts, or do you also catch slowdowns before the official page turns orange???