Hacker News Mention Monitor with Hermes
One rule overrides everything else here: a thread accusing you of a security problem or a data leak goes to the top with a plain line saying a person is needed now, whatever your ping policy says. Recognising that in a comment which never uses the word security is judgement, not keyword matching. Hermes brings general reasoning to that call on an open model foundation.
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.
Hourly sweep, Tuesday 14:00
The 14:00 tick fires. A Show HN posted 40 minutes ago is at 168 points and 74 comments, sitting fourth on the front page. The top comment, from user tkachenko, asks how self hosting compares with the incumbent, two people report that the docs link returns a 404, and a competitor's founder has replied twice in the thread.
Set up in minutes
Using this template drops you into a guided setup. It asks exactly this, nothing else:
Connect Hacker News
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
Every term worth watching, in your own words: your product and company names, your domain, your founders' Hacker News handles, and the competitors you want to hear about. The agent searches exactly these, so include the spellings and misspellings people actually type.
Ping policy
What earns an immediate Slack ping versus a line in tomorrow's roundup, for example a front page story or a complaint thread pings while a passing mention waits, plus the channel and the hours you are happy to be interrupted in.
Runs on Hermes
Preselected for this page — connect your Hermes 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 Hermes for this agent
Spots the serious thread
A four-hour-old comment about records reaching the wrong customer has to outrank whatever else the hour turned up. That is reading intent rather than matching terms.
Open model with real tools
The agent reasons on open weights and still reaches Hacker News search and your Slack channel through NoClick's connections.
Before you fork
Is an open model reliable enough to decide what wakes me up?
The decision is bounded by the policy you write, and the escalation rule is absolute rather than discretionary, so the model is applying your judgement more than its own. Security Claim At Night is the run to read, since it fires at 23:00 and the correct output interrupts you anyway. A false alarm costs you a minute, and the reverse costs considerably more.
Run it with a different agent
Put Hacker News Mention Monitor to work on Hermes
Free to start. Guided setup, a test run against staged conversations, and it's live.