Skip to content
Agents tracked: 502 Releases (7d): 613 Agent status: 4 with issuesSince last limit reset: Claude 7d · Codex 5d For AI agents: Markdown Updated Nov 18, 2026
Case study XBOWResults as reported by Clay

XBOW built its own prospecting tool on Clay CLI and reached 100% rep adoption

XBOW, which builds autonomous offensive security agents, dropped its legacy data provider and had its GTM engineering team build an in-house prospecting and enrichment platform on Clay's CLI and Agent Plugin. Reps ask for contacts in plain language, enrich them from a browser extension and get their top accounts in Slack; the tool reached 100% rep adoption, as reported by Clay.

The problem

XBOW's revenue stack had grown to about 15 tools, so answering one question about an account meant switching between systems. Sales leadership asked for one place where sellers could get everything they need to sell. The legacy data provider was up for renewal, and the GTM engineering lead works in code and had never used Clay's web interface, so a point-and-click tool was not the answer.

How they did it

  1. Install Clay's Agent Plugin in your coding agent - in Claude Code run /plugin marketplace add clay-run/agent-plugins and /plugin install clay@clay-plugins, then /clay:setup to sign in and pick a workspace (Codex has its own marketplace command). You need Editor or Admin access in the Clay workspace.
  2. Build your internal tool against the Clay CLI and API instead of a separate provider API, so search, enrichment waterfalls and access permissions come from Clay and its 250+ data and integration partners.
  3. Watch what your best sellers do in the old tool and turn that into backend logic behind a natural-language request - XBOW translates "who I want to find" into a Clay search by title and seniority, then calls enrichment for work email, title and company data.
  4. Add a browser extension step for profile pages - a pre-qualification check that combines internal data with outside research, and a "Send to Salesforce" button that enriches through Clay only after the seller decides.
  5. Deliver account scoring where reps already work - XBOW sends each rep's top three accounts to Slack and lets them ask why an account ranks high.
  6. Build security in from the start - strict access controls, tenant isolation, encryption, audit logging and safeguards for prospect data - and check credit use per run, since CLI calls spend the same Clay credits as work in the app.

From the official docs

clay login --device

Results as reported by Clay

  • 100% rep adoption and 90% month-over-month retention among XBOW's sales reps, as reported by Clay.
  • 30% more relevant, usable contacts per rep, as reported by Clay.
  • Reps who use the platform have a close-won rate 5–6% higher than those who don't, and in Q3 32 of the 36 sellers who closed deals used it, as reported by Clay.
  • New sellers without long experience or large networks create pipeline at a rate similar to experienced sellers, as reported by Clay.

These figures are as reported by Clay in Clay customer story ↗. Clay sells the product used here. AgentGid did not measure or independently verify them.

Takeaway. If your team already writes code, use the data platform as a back end through its CLI and put the seller-facing experience in the tools reps already open, like Slack and the browser.
AgentGid's take

This suits companies with a GTM engineer who can build and maintain an internal app; a sales team without one is better served by Clay's own web tables. Per AgentGid's Clay page (pricing checked Oct 14, 2026), Clay has a free plan with 500 actions and 100 data credits a month, Launch from $167 a month, Growth from $446 a month and custom Enterprise pricing. Clay's docs say the Agent Plugin costs nothing extra, but CLI calls spend the same credits as work in the app, and the plugin needs a paid Claude or ChatGPT plan for the coding agent. Caveats: this is Clay's own customer story with no publication date, and the close-won gap compares reps who chose to use the tool with those who did not, so it shows correlation, not cause. If you want to compare outbound tools first, our catalog lists 11x, AiSDR and Regie.ai as Clay alternatives.

The agent used here

Similar use cases

Guides