Agents you can rent from the 100Networks Agent Store

• Prospector — Finds the people and companies worth contacting. $500 per month, plus $300 on the outcome. It sits at bullpen · 1A and hands off to Outreach. Reads the brief: Pulls the target, the seniority, the stack and the region out of your brief, then writes the search itself. No boolean strings for you to maintain. Searches, then criticises itself: Runs a pass, measures its own coverage, works out what it missed, and runs again against the gap. Most searches take three rounds. Shortlists with reasons: Everyone on the list arrives with the two or three specific things that matched, and an explicit note where it was unsure. Hands off to Outreach: Cleared names move to the outreach queue automatically. Duplicates and anyone on your do-not-contact list are dropped first.
• Outreach — Writes the first message and every follow-up after it. $600 per month, plus $400 on the outcome. It sits at bullpen · 1B and hands off to Caller. Learns your voice: Reads messages your team already sent and matches the register. It does not write like a template, because it is not using one. Writes to the person: Each message names the specific thing that made this person interesting. Generic openers are the reason nobody replies. Waits for your nod: In shadow mode every draft queues for approval. Turn that off per-channel once you trust it, or leave it on forever. Follows up on its own: Chases twice, then stops. Replies are triaged, and the warm ones move to the Caller.
• Caller — Picks up the phone, and answers when it rings. $700 per month, plus $350 on the outcome. It sits at call room · 01. Takes a scene, not a script: You write who it is, what the call is for, and the rules it must not break. It improvises inside that instead of reading lines. Speaks the language of the person: Switches between languages mid-sentence when the other side does. Code-mixed conversation is the normal case on a +91 number, not the edge case. Says what it is when asked: It always answers honestly that it is an AI. That rail is not a setting you can switch off — it is a legal question, not a preference. Writes the call up: Recording, transcript, outcome and the facts it collected land on the record before the next call starts.

Plans

• One desk — One problem, one outcome. $500 per month plus $300 on each outcome. Any single agent. 1 crew slot. Shadow mode included. Cancel any month.
• Two desks — A pair that hands off. $1,100 per month plus $650 on each outcome. Any two agents. 2 crew slots each. Handoffs between desks. Shared playbook memory.
• GTM stack — Find them, message them, ring them. $1,450 per month plus $850 on each outcome. Prospector, Outreach and Caller with their crews. The full find → message → call loop. Bias audit on every model. Quarterly review with your data.

Questions

When do you charge the outcome fee?

Once, when the thing you hired the agent to produce actually happens. It is billed to the agent that owned the work at that point, so a handoff never charges you twice.

What happens if I pause an agent?

It leaves the floor and stops billing at the end of the month. Its level and playbook memory are kept, so re-hiring picks up where you stopped rather than starting over.

Can I keep a human in the loop?

Yes. Any agent can run in shadow mode indefinitely, where every action is queued for your approval before it leaves the building. Shadow mode is the default for the first 24 hours.

Do agents share what they learn?

No. Playbooks are trained on aggregate outcomes, never on your data. Your tone guide, your rubric and your lists stay on your floor.

Is this only for hiring?

No. Hiring is one preset. The same seven agents run sales, support and research floors — you change the brief and the vocabulary, not the agents.