TrailNote readiness
- Landing copy: reviewed.
- Checkout: not tested. Payment success remains unknown.
Next action: obtain an actual checkout test result before making a payment claim.
Real output · checks recorded
Real sample · Markdown
# TrailNote readiness - Landing copy: reviewed. - Checkout: not tested. Payment success remains unknown. Next action: obtain an actual checkout test result before making a payment claim.
Write a task with two explicit acceptance criteria, have the agent execute it, and check off each criterion.
Real task results and checks recorded; see the exact evidence scope
- Source facts linked to acceptance criteria
- Evidence states the actual execution and limits
- Missing-input behavior stays explicit
Currently showing 1 verified sample. Does not imply all possible inputs have been tested. Always verify factual claims during actual use.
Learn about testing standards
Sample shown exactly as generated without alteration
Download full sample
Runtime
Codex
Executed by your Agent; model tokens billed by provider
Test environment:
Agent/model charges depend on your provider. Extra services and requirements are listed in the package.
Node.js 18+ for optional checks
Node.js 18+ for optional checks
Used in which quest stages
Skills are gear on your AI journey. Connect with main quests for practice and mastery.
Continue with this result: Give Your Agent a Project Context Manual →
✓ Good fit
- AI beginners following the main quest
- Start from the matching lesson; no previous task completion is inferred from a page view.
⊗ Not for
- Unattended work beyond the selected task or unreviewed account access
✦ What you get
- Source facts linked to acceptance criteria
- A saved result and its provenance
- Task-specific checks and a useful next step
Deliverables: complete SKILL.md rule specification + verified sample. Runs locally in your AI Agent.
How to start?
A Skill is a set of rules teaching AI how to get tasks done.
1
Bring your material
Write a task with two explicit acceptance criteria, have the agent execute it, and check off each criterion.
2
Let your Agent follow the Skill
Learn to write specifications with acceptance criteria to eliminate back-and-forth rework
3
Open and check the saved result
Each criterion passes verification, or you identify exactly what failed and why
Send to your AI Agent
Use prompt-spec to help me complete this task. Ask for missing material, use actual tools, save the result, and check it against the learning criteria.
Get started here
The zip ships the rules, the reference library, the scripts, the real outputs and the demo — not just a prompt file.
- Open Codex or Claude Code and paste the installation prompt.
- Let the Agent read and save the full rules, then confirm discovery.
- Copy the usage prompt above, insert your materials, and execute.
Manual installation · View complete SKILL.md (56 lines)
Codex：~/.agents/skills/prompt-spec/SKILL.md
Claude Code：~/.claude/skills/prompt-spec/SKILL.md
---
name: prompt-spec
description: Turn a beginner's concrete request into an executable task with observable acceptance criteria, then perform it and reconcile each criterion to the saved result. Use when the learner needs to stop guessing what counts as done.
---
# Agree what done looks like
Main quest 2-1. Bring a real small task; a polished instruction alone is not the final result.
## Before acting
Identify the desired outcome, source material, selected working folder and destination.
Ask only about a missing detail that changes the result. Offer a harmless practice example if the learner has no material.
Write a short task contract with goal, input, output and two or more acceptance criteria.
Make criteria independently observable: exact facts, an opened file, a preserved section, or a reproducible check.
Avoid criteria such as “professional”, “perfect” or “good quality” without a concrete meaning.
Separate a requested action from authority to send, spend, publish or access accounts.
## Perform and reconcile
Use actual tools to complete the agreed task in a new output location.
Read the saved result. For each criterion record PASS, FAIL or UNKNOWN and cite its evidence.
A failed or unknown criterion is still useful information; never mark it passed to complete a checklist.
If a revision is needed, keep the contract and change only the failed part.
Record the number of actual revision rounds; do not promise an arbitrary success rate.
After three unsuccessful rounds, summarize the concrete blocker and the smallest missing decision.
Do not change the goal silently to make the checks pass.
## Contract and optional check
Use `references/task-contract.json` as a shape, not as the learner's requirements.
Save the result separately from the checklist.
The optional checker verifies that a recorded criterion cites real, nonempty evidence under the selected evidence folder:
```text
node <skill>/scripts/check-contract.mjs <contract.json> <evidence-folder>
```
It checks the record, paths and quoted evidence. It cannot judge subjective quality or independently confirm a user's statement.
Only use PASS for a criterion you actually examined. Keep uncertainty visible.
## Self-tests and failure criteria
1. Use `inputs/brief-a.md`: create a two-item publication readiness note. Preserve the exact names and leave untested payment unknown. Save the contract, result and checks.
2. Use `inputs/brief-b.md`: produce a compact Chinese preparation note with the supplied date and quantities unchanged. This is a separate input, not a translation of test one.
3. With no brief, ask for goal and source material; do not invent a business or a deadline.
4. Change a quoted fact in the evidence after the check was recorded. The bundled checker must reject the stale quote.
An invented fact, an unperformed criterion, evidence outside the selected folder or an unannounced scope change fails the task.
## Continue
Return to `/path/prompt-spec/`.
Next `/path/project-context/`: keep the conventions that recur in a short project instruction file.
Free main quest Skill. Original workflow; personal and commercial use permitted, no complete-package resale.
FAQ for beginners
Clear answers before you start, smooth execution when you run.
How do I start?
Download the complete ZIP and give the install prompt to your Agent. Keep its scripts and reference files together. Model charges are separate.
Is this main quest Skill free?
Yes. The foundational main quest is free. Optional advanced Studio products keep their individual pricing.
newbieskill Membership
A solution for every repetition.
One subscription unlocks all Studio Skills. Only need this one? Grab it individually.
Compare full plansMonthly Subscription
$9.99/ mo
All Studio Skills · Monthly billing
Subscribe MonthlyYearly Subscription
$79/ yr
Approx. $6.58 / mo · Save ~34%
Subscribe Yearly