Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add LaGrowthMachine/gtm-system --skill won-deal-icp-findergit clone --depth 1 https://github.com/LaGrowthMachine/gtm-systemWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/lagrowthmachine/gtm-system/won-deal-icp-finder)<a href="https://agentmods.dev/skills/lagrowthmachine/gtm-system/won-deal-icp-finder"><img src="https://agentmods.dev/badge/skills/lagrowthmachine/gtm-system/won-deal-icp-finder/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/lagrowthmachine/gtm-system/won-deal-icp-finder"><img src="https://agentmods.dev/badge/skills/lagrowthmachine/gtm-system/won-deal-icp-finder.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00220 | $0.04423 |
| Opus 5 | $0.00110 | $0.02211 |
| Sonnet 5 | $0.00044 | $0.00885 |
| Haiku 4.5 | $0.00022 | $0.00442 |
Grade A, and why
won-deal-icp-finder scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 162 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Won-Deal ICP Finder
Turns a deal dataset into a proven ideal customer profile — which companies generated the value, what they have in common, and which channel won them — then helps find more like them.
Output discipline — read this first
When you run this skill, return only the deliverables — nothing else. No preamble ("Let me…", "I'll start by…"), no narration of the steps, no restating these instructions, no closing pitch beyond the single step-4 note. Each step is one sentence plus its table or widget — no analysis essays, no editorializing about what the numbers "mean" or "signal." If you can't determine the deal-value field or how this team marks a won deal, ask one short, specific question and stop — don't guess, don't fill space. Otherwise: output the four deliverables and stop.
Authority — read this first
Everything you need is inline in this file. There is no taxonomy JSON to grep.
- The numbers — ranking deals by size, aggregating revenue per company, concentration, segment breakdowns, ranking acquisition sources by frequency — are produced by
scripts/analyze.py. Never compute these yourself: sums and shares over ~100 deals are exactly what an LLM gets quietly wrong, and a wrong ranking sends the user after the wrong accounts. Run the script; reason over its JSON. - The judgment — clustering companies into named ICP archetypes, reading the source ranking, deciding what to flag — is your job, using the rules below.
examples/sample-deals.jsonis a fictional dataset for a worked run.scripts/analyze.py --testis the self-test.
What it does
The job, in four moves:
- Pull and rank won deals from the last 12 months — selected by deal value, not by a closed-won status that may not exist in this CRM — with their companies, ranked by deal size.
- Locate acquisition source. Where the source lives varies by HubSpot setup — inspect a sample deal + its company + contact to find the right field (standard or custom), then read it for all deals.
- Cluster into ICP archetypes — 2–4 named, criteria-based company profiles, each with a one-click "find more like this" via
sales-nav-search-builder. - Rank the acquisition sources behind these big deals (top 5 + values), and — when there's no campaign-level detail — flag the blind spot.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 162 lines · 220 tokens per session scan A f027bb3ad576
won-deal-icp-finder is a skill published in the GitHub repository LaGrowthMachine/gtm-system (37 stars, last pushed 3d ago), licensed MIT. It adds 220 tokens to every session and 4,423 once invoked, about $0.0011 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
deepline-ads-audiences
Use this skill when building, enriching, auditing, or uploading B2B paid ads audiences to Google Customer Match, Meta/Facebook Custom Audiences, or LinkedIn Matched Audiences. Triggers on phrases like '/deepline-ads-audience', '/deepline-ads-audiences', 'upload this audience', 'create custom audiences', 'personal…
deepline-pre-research
Use when the user wants a last30days-style pre-research pass in Deepline: discover the critical public, private, CRM, workflow, social, and web data sources for a research/enrichment job; compare provider coverage; estimate Deepline credit cost; recommend the source plan before building or running the workflow; or…
niche-signal-discovery
Discover niche first-party signals that differentiate Closed Won vs Closed Lost accounts for ICP analysis. Use when the user provides won/lost customer domain lists and wants differential signals (website content, job listings, tech stack, maturity markers) to build account scoring models and prospecting criteria.…
deepline-engine
Build, publish, and verify a Deepline Play as a durable state machine over Customer DB tables, including a small paid pilot. Invoke this skill explicitly when the user asks to build an engine.
deepline-plays-review
Use this skill when a human needs to review a Deepline Play result and hand feedback, labels, or approval back to the agent for revision, evaluation, comparison, or bounded iteration. Triggers on “put this run in a Sheet,” “review these results,” “read my feedback,” “make this a standing rule,” “never regress on this…
deepline-plays
Use for Deepline GTM work that searches, enriches, scores, collects signals, or automates a workflow: find companies or people, enrich a CSV, find emails or LinkedIn, compare providers, build a waterfall, create a webhook or cron, or write a Play. For live information work, run a small heterogeneous experiment…