Bluesky Mention Monitor with Claude Code
The mistake this job invites is reading a thread's temperature off the one post your search term happened to match, when that post is usually a reply hanging off somebody else's root. Getting it wrong sends a working day at a joke and misses the warning underneath it. Claude Code pulls the whole conversation and the account behind it before forming a view, which is the reading that prevents both.
Watch it work before it's live
Run a staged conversation — no account needed. The agent handles it for real while a simulated world answers its tool calls; nothing touches real accounts, and nothing is actually sent.
has anyone here actually run this in production behind a corporate proxy
Posted 25 minutes ago, 18 likes and 9 replies, moving through a self hosting feed that a few thousand infrastructure people read daily. She maintains a Python client our own docs link to and carries 1,900 followers. Her post sits as a reply inside a thread about egress costs, and two people below her say they hit the same wall last month and gave up on it.
Set up in minutes
Using this template drops you into a guided setup. It asks exactly this, nothing else:
Connect BlueSky
One sign-in. The agent acts through your account, scoped to what this template uses.
Connect Slack
One sign-in. The agent acts through your account, scoped to what this template uses.
Watch terms
Everything the sweep should look for: the company and product names, the domain you own, the handles people tag instead of typing a name, and any competitor you want to hear about. Terms are searched literally, so add the shortened forms and the wrong spellings your community actually types.
Ping policy
When a post is worth interrupting a working day and when it belongs in the once daily line instead: the accounts or communities that always count, the kinds of question you want reaching you fast, your quiet hours, and the channel all of it lands in.
Runs on Claude Code
Preselected for this page — connect your Claude Code account during setup, or switch to NoClick's built-in models with one click.
Watch it handle a test run
A staged conversation against a simulated world — then it’s live.
Why Claude Code for this agent
Root first, matched post second
The reply is the native unit on this network. Fetching the thread a hit belongs to is what separates praise from a complaint written in nearly the same words.
Who, inferred and stated
A handle resolving to a company domain places somebody more reliably than a bio does. The message says who it thinks an account is and what that was inferred from, so you can disagree with it.
Before you fork
Our name is also a common word. What does that do to a sweep?
Collisions are the ordinary case rather than a corner case, and they are discarded before anything is counted or sent. The practical cost is a slightly longer run each hour, not a noisier channel. If it still feels heavy after a week, cut the term list back to your handle and your domain, which on this network catches most of what matters because people tag accounts more than they type names.
Run it with a different agent
Put Bluesky Mention Monitor to work on Claude Code
Free to start. Guided setup, a test run against staged conversations, and it's live.