agentleFS
JakeschincariolJakeschincariol/whale-skill/market-researcher/SKILL.md

market-researcher

Skill15 starsChanged 9 days agoMITView on GitHub

Market Researcher. Finds problems businesses already pay thousands to solve. Looks for money that is already leaving their accounts: vendors, agencies, outsourced services and salaried jobs that exist only because of one painful problem. Builds a table of the problem, who pays and what they pay today, with a source for every number or a clearly marked estimate with its working, then helps you pick the one to go after and sets up the board the other ten Whale skills share. Use when the user says "what should I build", "find me a problem", "market research", "validate this idea", "who would pay for this", "what do businesses pay for", or "set up the whale skill".

12 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use frontend-design.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • BlockSkill

    frontend-design

    Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

9 rules. 0 are common, 6 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 9 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Don't start with an idea. Start with a line on somebody's bill.

    line 16
  2. Very specific:

    Every command takes --file (default whale/whale.json).

    line 35
  3. Very specific:

    Ask for their name, business name, the email they sell from, a postal address (the law needs one in cold email), and who they would like to …

    line 39
  4. Distinctive:

    Look for money already moving: Search the web for:

    line 48
  5. Distinctive:

    Talk to someone: Before picking, tell the user to ask one person in that business: "What do you pay for this today?" Research finds the …

    line 59
  6. Distinctive:

    Never states a number it can't source.

    line 81
  7. Distinctive:

    Never reads anything behind a login, a paywall or someone's private books.

    line 82
  8. Very specific:

    Never tells the user the idea is validated.

    line 83
  9. Distinctive:

    Never invent a number, a result, a client or a source.

    line 90

“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

market-researcher

Don't start with an idea. Start with a line on somebody's bill. If a business already pays thousands a month for something, the demand is proven and so is the budget. This skill finds those lines.

${CLAUDE_SKILL_DIR} below is this skill's folder (where this SKILL.md lives).

python3 "${CLAUDE_SKILL_DIR}/board.py" init --name "You" --company "Your Company" \
    --email hello@yourcompany.example --address "100 Main St, Austin, TX 78701" \
    --product "Your Product" --does "Chases your late invoices automatically" \
    --value_line "It costs about a third of what you pay a collections service today."
python3 "${CLAUDE_SKILL_DIR}/problems.py" add --problem "Chasing late invoices" --who "Freight companies" \
    --pay 6400 --phrase "chase late invoices" --source https://example.com/collections-pricing
python3 "${CLAUDE_SKILL_DIR}/problems.py" add --problem "Scheduling field crews" --who "Landscapers" \
    --pay 2900 --estimate --basis "one office admin, half time, at the local median wage"
python3 "${CLAUDE_SKILL_DIR}/problems.py" rank
python3 "${CLAUDE_SKILL_DIR}/problems.py" pick --id chasing-late-invoices
python3 "${CLAUDE_SKILL_DIR}/board.py" board

Every command takes --file (default whale/whale.json).

First run

Ask for their name, business name, the email they sell from, a postal address (the law needs one in cold email), and who they would like to sell to. Don't guess any of these. board.py init writes them down. --target_customers (default 3) is how many companies must pay before the idea counts as validated: set it now, before there is any money to tempt anyone into moving it.

Steps

  1. Pick a kind of business. One the user knows or can reach: an old job, a family business, a trade they have worked with. It is much easier to sell to people whose world you know.
  2. Look for money already moving. Search the web for:
    • Vendors and agencies that solve one narrow problem, with public price pages ("collections service pricing", "after-hours answering service cost").
    • Job ads that exist only because of the problem. A posted salary is a number you can cite.
    • Industry forums and reviews, where owners complain about what they pay and why.
    • Public filings and trade press, for the bigger companies.
  3. Write it down with the number. problems.py add. A number needs --source (a link, or what someone told you and when). No source means --estimate --basis "the working", and it ranks below every sourced one. --phrase is how it reads in a pitch: "pay about $6,400 a month to ___".
  4. Rank and read it back. problems.py rank prints the table. Under $1,000 a month is flagged: it is rarely worth a decision maker's time.
  5. Talk to someone. Before picking, tell the user to ask one person in that business: "What do you pay for this today?" Research finds the problem. A person confirms it.
  6. Pick one. problems.py pick. Then /demo-builder shows the answer and /whale-hunter finds who decides.

What makes a good one

  • Money is already being spent, and you can say how much.
  • It repeats every month. A one-off fix is a project, not a product.
  • One person's budget covers it, so one person can say yes.
  • You could show it working in 2 minutes with sample data.

Example from the reel

The table: Chasing late invoices (freight companies, $6,400/mo), Answering after-hours calls (plumbing and HVAC, $3,800/mo), Filing insurance claims (auto body shops, $4,500/mo), Renewing contracts on time (security firms, $5,200/mo), Scheduling field crews (landscapers, $2,900/mo). The pick is the first. At North Pier Logistics, "Invoice collections service, $6,400.00" is a line on their monthly expenses. A real company's books are private, so that number only counts if they showed it to you. The whole story is whale.example.json beside this file (illustrative: every company, person and figure in it is fictional).

What it never does

  • Never states a number it can't source. An estimate is always labelled as one, with its working.
  • Never reads anything behind a login, a paywall or someone's private books.
  • Never tells the user the idea is validated. Only money does that: /pre-sale validated.

Consent and truth

  • Nothing is sent, booked or charged without the user's explicit yes. This skill only reads and writes notes.
  • Public business information only: a company's own site, public listings, job ads and filings, never a person's private life, and nothing behind a login.
  • Never invent a number, a result, a client or a source. Fill in only what is true and ask when something is missing.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.