GTM AgentsA free Claude Code plugin marketplace for revenue teams: 67 plugins bundling roughly 204 GTM agents, 244 skills, and 20 orchestrators for sales, marketing, customer success, and RevOps, all as markdown prompts under Apache-2.0.
Repo

GTM Agents

A free Claude Code plugin marketplace for revenue teams: 67 plugins bundling roughly 204 GTM agents, 244 skills, and 20 orchestrators for sales, marketing, customer success, and RevOps, all as markdown prompts under Apache-2.0.

Open source 392Markdown (Python for repo scripts)Apache-2.0since 2025
Free & open source
The take

The broadest free GTM prompt-pack for Claude Code, 67 plugins across sales, marketing, CS, and RevOps, all Apache-2.0. Two honest catches. Everything here is markdown instructions, so results depend entirely on the context and tools you give Claude, and the repo has mostly idled since one big Nov-2025 push while its README still advertises hour-savings and dollar figures its own agent counts contradict.

Great for

Claude Code users who want pre-written GTM slash commands for lead research, email sequences, content calendars, and pipeline audits, and want a structured prompt library instead of a blank file.

The catch

The plugins are prompts with no integrations behind them. No prospect database, enrichment, or CRM connectors ship, so left alone Claude scrapes or guesses. Read the skill files, and run a security check, before you trust output in front of a client.

Bottom line

A genuinely well-scaffolded prompt marketplace with real CI gates, worth cloning for the structure. Treat it as a starting template, not production tooling, and expect to wire in your own data and keep the skills current yourself, since the repo has gone quiet since April 2026.

At a glance
Install/plugin marketplace add gtmagents/gtm-agents, then /plugin install <name> (e.g. sales-prospecting); Claude Code only
LicenseApache-2.0 (permissive, patent grant, no copyleft)
LanguageMarkdown skill and agent files (Python only for repo-maintenance scripts)
Maintained byLaunchIQ AI, Inc. (gtmagents org), effectively one primary maintainer
Harnesses supportedClaude Code (plugin marketplace plus slash commands); skills follow the Agent Skills open standard, but install is Claude-Code-only
Watch out forIdle since April 2026, README claims and its own agent and skill counts do not match, prompts run unsandboxed in your context, no built-in data or CRM connectors
Feature checklist
Sales plugins (prospecting, pipeline, enablement, calls)Yes
Marketing plugins (content, email, SEO, social, paid)Yes
Customer success + RevOps pluginsYes
Multi-agent orchestrators20 workflow orchestrators for campaigns, ABM, nurture, launches, and forecasting
Agent Skills open-standard complianceMigrated to the standard in a Dec 2025 pass across all skills
CI validation (manifest, frontmatter, smoke test)validate_marketplace.py plus a smoke test run on every push; eight more validators exist but are not wired into CI
Model-mix cost controlRoughly 60% Haiku by design, but the auditor script is not a merge gate
Tagged releases / versioningCHANGELOG.md exists, but no GitHub releases or version tags
Built-in data enrichment / prospect databasePrompts only; Claude scrapes or guesses without your own tools
CRM / tool integrationsNo native connectors; wire in your own MCP servers or APIs
Output-quality testsCI checks file structure, not whether prompts produce good output
Opus-tier agentsArchitecture doc claims 28 Opus agents; zero on disk, validator allows only Haiku and Sonnet

What it does

  • Installs as a Claude Code plugin marketplace: run /plugin marketplace add gtmagents/gtm-agents, then pick from 67 plugins and invoke them as slash commands
  • Bundles roughly 204 subagents and 244 skills across sales prospecting, pipeline, enablement, content, email, SEO, social, paid media, customer success, and RevOps
  • Ships 20 multi-agent orchestrators that chain steps for campaigns, ABM, lead nurture, product launches, and revenue forecasting
  • Structures every skill on Anthropic's Agent Skills progressive-disclosure convention, loading full templates only when a task actively needs them
  • Wraps 700-plus markdown files in a package-registry-style pipeline: a marketplace.json manifest, a frontmatter validator, a smoke test, and a model-mix auditor
  • Pins each agent to a Haiku or Sonnet model via an allowlist the validator enforces, keeping the roster roughly 60% Haiku for cost
  • Includes beginner docs (GETTING_STARTED, QUICK_START), eight use-case guides, recipe cards, and a Slack community for support

What it can’t do

  • Do any GTM work on its own; the plugins are markdown prompts, so every result depends on the context, data, and tools you give Claude Code
  • Enrich, dedupe, or pull leads from a real database; there is no built-in enrichment, prospect index, or CRM connector, so Claude scrapes or guesses unless you wire in your own
  • Guarantee output quality; there is no automated test that an agent's instructions actually produce good results, only file-structure checks
  • Deliver the Opus tier its own architecture doc describes; an independent review found zero Opus agents on disk (122 Haiku, 82 Sonnet), since the validator allows only Haiku and Sonnet
  • Be trusted from the README's numbers; the doc advertises agent and skill counts that contradict each other and the files on disk
  • Promise upkeep; after one big Nov-2025 build the repo has mostly idled, with a single skill tweak merged in April 2026 and 98 open issues

Pros & cons

Pros
  • Broadest free GTM prompt coverage in one Apache-2.0 marketplace, from prospecting and pipeline to content, SEO, CS, and RevOps
  • Permissive Apache-2.0 license with a patent grant, no copyleft or network-use obligations for commercial or client work
  • Unusually well-scaffolded for a prompt repo: manifest schema, frontmatter validator, smoke test, and a model-mix auditor run as CI gates
  • Every skill follows Anthropic's Agent Skills standard, so it plugs into Claude Code cleanly and reads as intended prompt design
  • Free on top of a Claude Code subscription, with no API key or signup to try the prompts
  • Model choices favor Haiku for most agents, keeping token cost down when you run the workflows at volume
Cons
  • Everything is a prompt library with no integrations; no data, enrichment, or CRM connectors ship, so real workflows need you to add your own tools
  • Repo has mostly idled since April 2026 after one big initial push, with 98 open issues and no tagged release
  • README makes inflated, unverifiable claims (15-plus hours saved per week, $25,000-plus per user per year, $2.3M ARR) and uses placeholder testimonials with generic names
  • Its own counts do not reconcile: 92 versus 203 agents, 52 versus 243 skills across the same README, none matching the roughly 204 files on disk
  • No output-quality tests; CI only confirms files exist and start with frontmatter, not that any agent produces useful work
  • Skill and agent instructions run in Claude's context with no sandboxing, so read the files and security-check the repo before client use
What 3 real users said
1 positive1 mixed1 critical

What the community says

Someone treated 700+ markdown files like a package registry with CI gates instead of a folder of prompts. That's the part worth reading, not the prompts themselves.
Florian NarrCodeline · May 14, 2026 ↗
The Opus tier described in the project's own architecture doc doesn't exist in the codebase, and couldn't, since the validator would reject it.
Florian NarrCodeline · May 14, 2026 ↗
If you want a Claude Code plugin marketplace with sales and marketing prompts pre-written, this saves you the scaffolding. If you're evaluating it as production tooling, read scripts/validate_marketplace.py before you read the agent prompts, it's the most rigorous file in the repo, and it quietly proves the README's own numbers are wrong.
Florian NarrCodeline · May 14, 2026 ↗
6 sourcesWhere this page’s facts came from. Open the receipts.
  1. 1GitHub repo (gtmagents/gtm-agents)
  2. 2GitHub API, repo metadata (392 stars, Apache-2.0, created Nov 2025, last push Apr 2026, 98 open issues)
  3. 3GTM Agents README (67 plugins, install, use cases, and the contradictory agent and skill counts)
  4. 4SECURITY.md (maintained by the GTM Agents team at LaunchIQ AI, [email protected])
  5. 5Codeline repo review by Florian Narr, disk-verified agent and model counts and the idle-since-April finding
  6. 6Claude Code Marketplaces directory listing for gtmagents/gtm-agents