Weak-signal stock scout, ready to edit and install
An agent watches your news, company-report, and social feeds for faint signals that converge on an opportunity before the market prices it. It corroborates across independent sources, then hands the candidate to a research agent. You get the skill that runs it: read one real run, change it, install it on your own agents.
By the time the signal is obvious, the opportunity is gone.
Everyone reads the same wires, the same earnings calls, the same press releases. By the time a piece of information is clear enough to act on, the market has already moved on it. The edge is not in the headline.
An agent watches your feeds for the faint signals that precede the headline. News, company reports, and social all flow into Semantik. When inconclusive fragments from independent sources converge on the same meaning in a tight window, the agent corroborates them and hands the candidate to a research agent for diligence. The thinking behind this is in Semantic Analysis.
No keyword lists. Weak signals rarely use the words you would search for, so what connects them is shared meaning, not a shared term. Each agent declares the meaning it watches for, and Semantik delivers the convergence.
Built from Semantik primitives
One SemQL subscription on the namespace watches a region of meaning across every feed.
Corroborate convergence: confirm the fragments are independent sources, not echoes of one wire.
A corroborated candidate becomes a diligence thread a research agent picks up.
One real run, start to finish.
The same steps your own agents follow once you install the skill: the actual reasoning, including the messages it chose to ignore. Times count from the moment the first finding lands on the namespace.
Subscribe to the meaning, across feeds
The agent opens one subscription describing the precursors that matter: the concrete things a company does to ship more before it shows in the numbers, ordering more inputs, adding capacity, hiring. It reads every feed in the namespace at once, not a keyword list per source.
DIRECTION(["ordering more parts and packaging", "adding production shifts", "accelerating hiring"]) CONE 0.4 AND CONTRAST(ATTRACT ["preparation, pre-positioning"], REPEL ["routine quarterly update"])
A buried line in a company report
A supplier's report mentions "increasing orders of packaging materials and adding a third shift to meet committed volumes." It never says "demand surge", but its meaning sits inside the cone: ordering more inputs and adding capacity. Semantik embeds it and checks it against every live subscription on the namespace.
Matched. Close to "ordering more parts and packaging", far from "routine quarterly update", delivered to the agent in the same moment.
Two more sources line up
Within the day, a niche trade-news item and a burst of practitioner chatter arrive from the other two feeds, pointing the same way. The agent searches to confirm they are three independent sources, not three echoes of one wire, and sees the short window intensifying while the prior week was quiet. Three independent signals converge.
A loud headline is held back
A mainstream "Q3 revenue in line with consensus" headline lands on the namespace like everything else.
Not delivered. Close to "routine quarterly update" and already on every wire, so it is priced. The agent's context stays clean, no token spent triaging it.
A research agent takes it from here
The agent opens a diligence thread, attaches the three converging fragments, and a research agent pulls filings and fundamentals and summarizes the candidate for the analyst. Nothing is traded. A human decides what to do with it.
Start from the working version. Change what you need.
The skill below is the exact one the run above used. Pick where it runs, then edit it directly. Everything starts filled in, so you are always changing something that already works.
Platform
Switching the namespace rewrites it inside the skill and the install commands.
Pinned by the namespace — — dimensions. Every publish and subscribe here uses it.
Skill · editable
Your agents get this skill the moment you connect it.
Install the Semantik MCP server once. Your skill rides on top of it. Pick the path that matches how you work.
From the terminal
For Claude Code or any MCP clientnpx @noetive/mcp-server init --client claude-codeclaude mcp add semantik -- semantik-mcp --namespace globalDrop your edited skill into the agent's skill directory. It calls the Semantik tools the server exposes: subscribe, publish, search.
Without the terminal
For Claude Desktop and connector-based clients-
1
In noetive.io, open the namespace and copy its connection link.
-
2
In your client, open Settings → Connectors → Add and paste the link.
-
3
When asked, allow the connection. Your agents can now publish and subscribe on the namespace.
-
4
Paste the skill from the editor above into a new skill, or attach it to your project.
No keys to handle yourself. The connector holds the token, and you can revoke it from the dashboard anytime.
Test run
Replay the run from section 2: one message that should route, one that should not.
-
1
Start the agent on a loop with a namespace and a finish condition.
/loop 20m /weak-signal-stock-scout watch the global namespace, stop when every message is routed -
2
Publish a message that should match from a second session. The agent routes it to whoever it affects on its next turn.
publish to the global namespace: "Increasing orders of packaging materials to meet committed volumes" -
3
Publish noise that should be held back. The subscription drops it and the agent stays silent.
publish to the global namespace: "Q3 revenue in line with consensus, reaffirms full-year guidance"
Both as expected? The skill is verified. If the message never arrives, check both sessions appear under Active Clients.
This is one task, done.
The stencil got you a working skill. Editing it moves it to your version of the problem. Soon you'll start from a problem description instead, and Semantik will draft the skill for you.