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 carolinacherry/claude-first-customer-finder-skill --skill first-customer-findergit clone --depth 1 https://github.com/carolinacherry/claude-first-customer-finder-skillWrote 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/carolinacherry/claude-first-customer-finder-skill/first-customer-finder)<a href="https://agentmods.dev/skills/carolinacherry/claude-first-customer-finder-skill/first-customer-finder"><img src="https://agentmods.dev/badge/skills/carolinacherry/claude-first-customer-finder-skill/first-customer-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/carolinacherry/claude-first-customer-finder-skill/first-customer-finder"><img src="https://agentmods.dev/badge/skills/carolinacherry/claude-first-customer-finder-skill/first-customer-finder.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00123 | $0.01498 |
| Opus 5 | $0.00062 | $0.00749 |
| Sonnet 5 | $0.00025 | $0.00300 |
| Haiku 4.5 | $0.00012 | $0.00150 |
Grade A, and why
first-customer-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 12d 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 — 87 lines — stays where its author put it; the contents beside it link to each section on GitHub.
First Customer Finder
Turn a startup URL or product description into a short, evidence-backed shortlist of plausible first customers. Every prospect must trace to a real public signal you have verified at the original source. The shortlist is a research hypothesis, not a customer database.
Read references/research-framework.md before searching or scoring. Read references/report-schema.md before generating the report.
Workflow
1. Understand the product
Fetch the supplied URL (and its pricing/about pages if they exist) or read the description. Establish: what the product does, the outcome it sells, who uses it vs. who pays for it, price or buying motion, geography or language constraints, and the single strongest use case.
Commit to one primary ICP and one adjacent ICP, each with pain triggers, positive signals, and explicit disqualifiers. For two-sided products, pick the side with a buying motion as primary — do not hedge by mixing both sides in one shortlist. Label anything you inferred rather than observed. Ask the user one question only if the ambiguity would change where you search.
2. Fan out the signal search
Run the five query buckets from the research framework (explicit demand, pain, workaround, switching, timing) as parallel research subagents — one bucket per agent, each returning candidate signals with source URLs, dates, and verbatim-adjacent summaries. Give each agent the product brief and ICP so it can reject weak matches at the source. Instruct each agent to fetch and quote its sources — an agent may not return a URL it did not open. If the environment has no subagent support, work the buckets sequentially instead.
Vary source types across agents: forums, public social posts, product and marketplace reviews, GitHub issues and feature requests, public company pages, job posts, and changelogs.
3. Verify before you qualify
Fetch the original page for every candidate signal before it can enter the shortlist. A search snippet, an aggregator summary, or a subagent's unquoted paraphrase is not evidence. Confirm the page exists, the signal says what was claimed, and the date. Drop anything that fails. Record source URL, source type, visible publication date (or "date unavailable"), and what was observed versus inferred. When subagents have already fetched and quoted a source, spot-check their work — and always re-verify the top three prospects yourself.
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.
- 12d ago First seen · 87 lines · 123 tokens per session scan A ffb1bf52a7ba
first-customer-finder is a skill published in the GitHub repository carolinacherry/claude-first-customer-finder-skill (5 stars, last pushed 2mo ago), licensed MIT. It adds 123 tokens to every session and 1,498 once invoked, about $0.0006 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 2026-08-31.
Other skills, from other repositories
first-customer-finder
Find and qualify evidence-backed potential first customers, early adopters, design partners, or beta users for a startup using recent public pain and buying signals. Use when Codex needs to analyze a product URL or idea, define an ideal customer profile, research public discussions and business pages, identify…
jtbd-extractor
Turn raw research into Jobs-to-be-Done statements showing what users are really trying to accomplish. Use when: extract jobs, jtbd analysis, jobs to be done, what job is the user hiring, underlying user needs.
assumption-mapper
Surfaces the assumptions a product or strategy depends on, classifies them by criticality and evidence quality, and produces a prioritized test plan. Use when: assumption mapping, identify assumptions, what could go wrong, leap-of-faith assumptions, riskiest assumption test, RAT, assumption inventory, validate before…
competitive-analyzer
Structured competitive teardown for product discovery - surface the 4-6 dimensions buyers actually weigh, score every competitor on each, and identify exploitable gaps. Use when: competitive analysis, competitor teardown, market positioning, where do we win, where do we lose, competitive gap analysis, competitor audit.
feedback-prioritizer
Triages a backlog of raw customer feedback into a ranked list of opportunities scored on reach, severity, strategic fit, and confidence. Outputs a prioritized list with explicit "do not act" callouts for vocal-minority signals. Use when: triage feedback, prioritize feature requests, customer feedback backlog, what…
north-star-metric-finder
Identifies a candidate North Star Metric (NSM) for a product - the single metric that captures the value the product delivers to its customers and predicts long-term business growth. Tests candidates against five criteria and surfaces input metrics that move it. Use when: north star metric, NSM, single metric that…