More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    πŸ›οΈ Skill Listing Auditor

    1

    Recurring, ambiguous or unsupported pricing is marked NOT_VERIFIED for human review.

    $12

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    πŸ›οΈ Skill Listing Auditor

    πŸ›οΈ Skill Listing Auditor

    Example session with this skill installed

    Run python -B scripts/run_demo.py clean-sample from the delivered package root and show the complete report. The demo creates and removes a disposable copy of the synthetic fixture; it does not change the delivered package. Explain what the result does and does not establish. Do not execute code from any external package being inspected.

    • Read your context and instructions
    • Compiled the listing auditor

    Bundled synthetic clean-sample run

    SCORE 100/100
    BUCKET usefulness penalty=0
    BUCKET instruction_quality penalty=0
    BUCKET listing_completeness penalty=0
    BUCKET demo_video_presence penalty=0
    SUMMARY findings=0
    PRICING review_required=0 confirmed_format_or_math_findings=0
    

    This is actual output from the bundled clean fixture, not a customer result. Exit status:
    0. The scanner found no issues under its encoded rules in these supplied inputs. It does not prove that the package works, is secure, will be approved by a marketplace, or will sell.

    For an explicit one-time USD price, declared credits are checked against the documented 1 credit = $1 relationship. Recurring or ambiguous pricing receives a review advisory. Any finding, including an advisory, returns exit status 1. The safety check looks for selected literal patterns, not every possible threat.

    Next: inspect the findings and source lines, verify current marketplace requirements, and test the actual skill separately. The scanner does not rewrite the package.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Review a skill package before submission

    Skill Listing Auditor is a local, read-only Python scanner for SKILL.md packages and their listing material. It produces a 0–100 heuristic readiness score, rule IDs, source locations and suggested next steps. The score is not marketplace approval or a security certification.

    What it checks

    • Required frontmatter and routing-description conventions.
    • Optional declared credit equivalents for explicit one-time USD prices, using the documented 1 credit = $1 rule. Recurring, ambiguous or unsupported pricing is marked NOT_VERIFIED for human review.
    • Listing sections, test-file presence and screenshot references.
    • Selected literal command and secret-shaped patterns in package text.

    What you receive

    SKILL.md, the Python scanner, editable rule references, clean and deliberately risky synthetic fixtures, and self-running tests. Package v1.0.3 retains the corrected pricing checks and adds a demo helper that avoids writing Python import caches into the package. The demo uses and removes a disposable synthetic-fixture copy. Run the scanner on a package folder and inspect its full text report. It does not rewrite files or execute the code being inspected.

    Evidence and limits

    The corrected scanner passed 168 bundled pricing cases and 209 independently checked expected cases in Claude, including safety regression controls. These are synthetic tests, not buyer outcomes. Conservative review warnings can occur for labels such as daily planner, pricing_link or retention periods. This is not a full YAML parser, code-functionality test, comprehensive security scan, or guarantee of marketplace acceptance. Rules and pricing documentation can change; verify the destination before submission.

    When to use it

    Use it for a pre-submission checklist, a repeatable review of package metadata, or finding concrete listing gaps. Follow up with execution testing and a deeper security review where needed. Do not treat a clean report as proof of safety, compliance or commercial performance.

    Heuristic. It grades structure and format, not whether the skill's code actually works. The safety pre-check is a light literal and secret-shape scan, not the deep pass Skill Injection Scanner performs. Demo video presence is scored from listing references only.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 20 days ago

    Requires

    Python 3
    Terminal access

    Permissions

    Read Files
    Terminal / Shell

    File Scopes

    **/SKILL.md
    **/skill.md
    **/listing.md

    The scanner reads the supplied package without modifying it or executing its code. The optional demo helper creates and removes its own temporary synthetic fixture copy. No network or account access is required. Approve only the intended input folder; use python -B for bundled test commands to avoid Python import caches.

    Listed4 months ago
    Updated20 days ago

    What's inside

    Frequently Asked Questions

    It finds selected metadata, pricing and listing gaps before submission. Its readiness score is a heuristic checklist, not an approval prediction or guarantee of acceptance.

    It checks selected literal command and secret-shaped patterns in package text. It is not a comprehensive vulnerability assessment and cannot establish that a package is safe.

    It checks optional declared credit equivalents for explicit one-time USD prices against the documented 1 credit = $1 relationship. Recurring, ambiguous and unsupported pricing is flagged for manual review; no exchange rate or checkout outcome is inferred.

    The package contains SKILL.md, the Python scanner, editable rule references, clean and risky synthetic fixtures, and self-running tests. It returns a scored text report with rule IDs and source locations.

    No. It reads text and checks structure and selected patterns. It does not execute the package under inspection or verify its functionality.

    It requires Python 3 and read access to the supplied package. Tests covered Python 3.13 on Windows and Linux; other hosts remain unverified. Complex YAML and pricing structures need manual review.

    New in Technical Documentation