interactive tool free members get 1 full audit

Rate your custom skills before you ship them.

Paste a SKILL.md draft and get a hard score on trigger clarity, workflow depth, safety, verification, portability, and buyer value. When API keys are configured, Boostor cross-checks the local audit with Codex and Claude.

If a skill can't say when it should stop, it is not ready to sell.

Structure— frontmatter, triggers, steps

Safety— stop rules, secrets, shell risk

Proof— tests, outputs, receipts

Cross-check— Codex + Claude when keys exist

> skill audit

Score the skill contract.

The widget treats the pasted skill as untrusted text. It never executes commands from the skill body.

skill input

Paste a SKILL.md draft.

target runtime

buyer / user

skill markdown

---
name: launch-readiness-review
description: Use this skill when the user asks whether a web app is ready to launch, especially before checkout, auth, analytics, or deploy changes go live.
---

# Launch Readiness Review

## When to use this
Use this when a user asks for a launch audit, release gate, production readiness check, or "does everything work?" review. Do not use it for small copy edits or purely visual polish unless the user asks for launch proof.

## Inputs
1. Repo path and branch.
2. Target URL if hosted proof exists.
3. Payment/auth/analytics surfaces that must be checked.

## Instructions
1. Verify the branch and worktree before making claims.
2. Inspect checkout, auth, analytics, env, and deployment configuration.
3. Run local lint/build checks when available.
4. Smoke the core route or hosted page if the user provided a URL.
5. Separate local proof, hosted proof, and payment-readiness proof.
6. Stop and ask before triggering real payments or destructive deploy actions.

## Verification
Report exact commands, URLs, files inspected, failures, and remaining blockers.

## Examples
- "Audit this branch before launch" -> produce a release-gate report.
- "Is checkout ready?" -> inspect checkout config and state what is verified vs blocked.

Free members get 1 full AI cross-check. The local structural score still runs when model keys or quota are unavailable.

audit report

Audit result appears here.

Paste a skill, choose the runtime, and run the rating pipeline.