A codified stack inverts that. Every tool in it can be driven headlessly, through an API, CLI, MCP server, or webhooks, which means the workflow lives in code an agent can execute, inspect, and repeat. The logic stops being tribal knowledge and becomes something you can read.
This page is the subset of our directory that clears that bar. Not a marketplace, not a reseller catalog: the tools we would wire into a code-run GTM system, each with the reason it qualifies.
A code-run GTM motion still follows the same pipeline as a manual one. The difference is what executes each stage. Here is where each qualifying tool sits.
BetterContact
Offers an API in addition to its CSV upload UI, so an agent can call it programmatically.
Clay
Clay exposes HTTP API and webhook steps so an agent can trigger enrichment runs, but the core workflow logic still lives in its visual table builder.
In our stackFree tier
Findymail
Provides a public API for email/phone lookup in addition to its web UI.
FullEnrich
Has a public API for bulk enrichment that an agent can call directly, alongside a simple UI for manual uploads.
In our stackFree tier
LeadMagic
Purpose-built REST API with a public playground and docs, no meaningful UI-only path.
MoltSets
Public API + MCP server (developer.moltsets.com); 20+ vendor waterfall behind flat-rate endpoints. BlitzAPI competitor.
Free tier
Prospeo
Has a documented public API for programmatic lookups alongside its browser extension.
Free tier
Apify
Every Actor is callable via REST API or CLI and returns structured JSON, so an agent can run and poll scrapers headlessly.
In our stackFree tierBlitzAPI
HTTP API at api.blitz-api.ai; LLP already drives it from Claude Code with per-directory keys.
In our stackExa.ai
Pure REST API designed for programmatic/agent calls, no UI required to operate it.
In our stackFree tierFirecrawl
Headless REST API purpose-built for AI apps and agents; no UI needed to run it.
In our stackFree tier
EmailGuard
It exposes an API for placement tests and DNS/blacklist checks that can be called programmatically from a sending pipeline.
Resend
Full REST API and SDKs let an agent create domains, send emails, and manage webhooks headlessly; this is already in LLP's own stack for transactional email.
In our stackFree tier
Our orchestration layer is Claude Code. Research that used to live in a browser tab runs as agent sessions with file access and judgment rules. Enrichment runs as per-task waterfalls in Clay, with the vendor order and credit cost tuned per job rather than one fixed cascade, and BlitzAPI inside the waterfall for prospecting data.
Sends go out through SmartLead over its API. Visitor and intent signals come back in from RB2B by webhook. The point is not any single tool. It is that every step can be triggered, checked, and corrected from code, so the judgment stays inspectable.
One honest boundary: codified does not mean UI-free everywhere. Clay qualifies because we drive it programmatically, but its core workflow logic still lives in a visual table builder, and we say so on its page. The badge marks what an agent can operate, not a purity test.
Full directory, including the tools that run on clicks: every tool, by category.
The criteria behind every badge on this site
Every tool here carries labels: a tier, best-for chips, an ease rating. Labels are cheap. So here is exactly what each one means and what a tool has to do to earn it.
- In our stack
- We run it on live client work today. Anything with this badge has a "How we run it" section written from actual use, not a trial.
- Recommended
- We evaluated it and would point a client to it in the right situation, but it isn't part of our daily stack. No hands-on usage claims appear on these pages.
- Previously used
- We ran it, then migrated away. The page says why.
Ease is judged for the majority buyer of the tool, and always ships with a one-line reason specific to that tool.
- Easy
- Self-serve. Productive in an afternoon.
- Medium
- Real configuration. Expect about a week of iteration before it earns its keep.
- Hard
- Needs an operator or engineering discipline. We say so plainly, because pretending a hard tool is easy helps nobody.
Best-for chips are capped at two per tool. A tool tagged for everyone is a classification failure, so most carry one.
- Agencies
- Repeatable across clients: workspaces or sub-accounts for client isolation, and pricing that does not punish usage or client count. Flat or per-seat beats usage-metered.
- SMB
- Runnable without a dedicated ops person. Fast time to value, simple and forgiving pricing. SMB problems are different problems, not smaller enterprise ones.
- Enterprise
- Admin controls, permissions and SSO, security review readiness, team scale. Enterprise ops teams absorb complexity, so tech-heavy is acceptable here.
- AI-native
- Built for builder-first teams that prefer API-first, composable, scriptable tools. Docs and API quality matter more than UI polish. Correlates with the Agent-ready badge by design.
- B2B SaaS
- An in-house GTM team building its own motion: native CRM integrations, product-data hooks, per-seat pricing works fine.
Agent-ready means the tool can be driven headlessly, through an API, CLI, MCP server, or webhooks, so a coding agent can operate it without a human clicking through a UI. It is the backbone of a code-run GTM system. The qualifying subset, assembled as a system: The Codified Stack.
You will not find a 4.7 out of 5 here. Scores imply a measurement methodology nobody publishes. We give you a verdict in sentences you can disagree with instead.
Every page shows when it was last reviewed. Pricing is verified against the vendor’s own pricing page as of that date, and corrected when we or readers catch drift.
Some links pay us a commission. That never changes what we list or what the verdict says, and each paid link is disclosed right where it appears.