agentleFS
mardab96mardab96/landing-pages-claude-skills/trust-signal-audit/SKILL.md

trust-signal-audit

Skill2 starsChanged 4 months agoMITView on GitHub

Reviews whether the page answers the questions a cautious buyer asks before handing over data or money: who runs this, what happens to my data, what if it goes wrong. Use when a form or checkout is about to ship, or when form abandonment is high on an otherwise clear page.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 5 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the page's ask cannot be identified, ask which conversion is being reviewed.

    line 23
  2. Very specific:

    Not for: judging testimonials and proof quality (social-proof-strength-audit), or form mechanics (form-friction-finder).

    line 25
  3. Distinctive:

    Do not describe a missing policy as a conversion percentage.

    line 41
  4. Distinctive:

    Do not give legal or compliance advice, and do not assess whether a policy is adequate.

    line 65
  5. Distinctive:

    Do not draft policy text. Say what question needs answering.

    line 66
  6. Distinctive:

    Do not convert a missing trust signal into a predicted lift.

    line 67

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “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

Trust Signal Audit

Use this skill when

  • the page asks for personal data, payment or a call, and the copy around that ask is available for review.
  • the page asks for card details, personal data or a document upload.
  • abandonment concentrates at the payment or submit step rather than earlier.

This skill diagnoses from absence, at the moment of the ask. It does not assess whether any policy is legally sufficient.

Required input

  • what the page asks for: personal data, payment details, a call, a document upload.
  • the copy immediately around that ask, which is the only place these signals do work.
  • what the page says about the company: legal entity, address, who is behind it, how to reach a human.
  • terms the buyer needs before committing: refund policy, cancellation, data handling, contract length, what happens after they submit.
  • optional: abandonment rate at the ask, and where in the flow it happens.
  • If the page's ask cannot be identified, ask which conversion is being reviewed. Trust signals are only assessable relative to a specific commitment.

Not for: judging testimonials and proof quality (social-proof-strength-audit), or form mechanics (form-friction-finder).

Analysis workflow

  1. Inventory trust signals: company info, privacy, security, reviews, proof, guarantees, certifications and contact routes.
  2. Match trust signals to the risk the visitor feels at the conversion point.
  3. Identify missing trust signals for sensitive forms, high-consideration offers or technical claims.
  4. Flag trust signals that are generic, buried or disconnected from the main objection.
  5. Recommend trust additions or placement changes near decision points.

Decision rules

  • Position is the whole finding. A refund policy in the footer does not answer a hesitation that happens at the card field. Report distance from the ask, not presence on the site.
  • The bigger the commitment, the more the absence costs. A missing legal entity on a newsletter signup is nothing; on a checkout it is disqualifying.
  • A contact form as the only support route signals that nobody will answer quickly. Name it, because first-time buyers weigh it heavily.
  • Trust defects are diagnosed from absence, not from performance data. You can state that a needed answer is missing without any conversion number, and you should say that is how you concluded it.
  • Do not describe a missing policy as a conversion percentage. Say which question goes unanswered at which moment.
  • Compliance and legal adequacy are out of scope. Whether a policy exists and is reachable at the ask is in scope; whether it is legally sufficient is not.

Output format

Question a cautious buyer asksAnswered on the pageDistance from the askCommitment at stakeSeverity
Who runs this / what happens to my data / what if it goes wrong / when will someone contact me / can I get my money backYes / no / only in the footer / only in a linked policyAdjacent / same screen / elsewhere on the site / nowhereThe specific askDisqualifying / costly / minor

End with:

  • Unanswered at the moment of the ask: the list
  • Decision: move terms next to the ask / add a real support route / name the legal entity / ask for data
  • Out of scope: legal sufficiency of any policy
  • Missing data: abandonment position, if it would change the ranking

Practical example

User: "Checkout page. No refund policy visible, no company name or address, privacy policy in the footer only, no security indicators near the card fields, support contact is a form. Cart abandonment at payment step is 71 percent."

Assistant should note: at the moment a card number is typed the page shows no refund terms, no legal entity and no support channel with a human on the end, so every reason to hesitate is unanswered exactly where hesitation happens; a privacy policy reachable only from the footer is not an answer at the point of the ask; a contact form as the only support route signals nobody will answer quickly, which matters most to a first-time buyer. Then: 71 percent abandonment at the payment step is where the money leaves, and the fix is placing refund terms, entity identity and a real support route adjacent to the card fields rather than elsewhere on the site. Whether these lift completion is a hypothesis to measure; trust defects are diagnosed from absence, not from performance data.

Guardrails

  • Do not give legal or compliance advice, and do not assess whether a policy is adequate.
  • Do not draft policy text. Say what question needs answering.
  • Do not convert a missing trust signal into a predicted lift.
  • Say plainly that these findings come from absence rather than from performance data.

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.