Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    GourmetSkill

    1

    Review your AI skill listing and price across eight dimensions. Get evidence-linked scores, prioritized fixes, and ready-to-use copy—with uncertainty made explicit.

    $6.89

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    GourmetSkill

    GourmetSkill

    Example session with this skill installed

    Use GourmetSkill to review the complete fictional listing below and its price. Respond in English with the listing diagnostic, all eight weighted dimensions, section comments, prioritized fixes, and ready-to-use replacement copy.

    Judge the platform purchase information alongside the description. Do not test the underlying skill or predict commercial outcomes.

    After the main review, run this separate access-limit check: suppose the entire PRICE AND PLATFORM INFORMATION section was unavailable to you. Explain what changes in scoring eligibility without treating unseen information as missing for buyers.

    Demo conditions: fictional product and platform terms; complete supplied-text review for the main case. No live website, real customer, or verified purchase is involved.

    PRODUCT
    ResearchBrief

    SUMMARY
    Transform your research into a clear, structured brief.

    ABOUT
    For researchers combining notes from papers. Start with source notes and a research question; request a 400-word brief, a section outline, a source list, and open questions. The skill is supplied as access through your connected AI assistant.

    ResearchBrief is designed to organize notes, suggest a narrative, and flag claims that need support. It can also check sources online. The example below shows how the brief format works.

    HOW TO START
    Load ResearchBrief in Claude or Codex, paste your notes, and ask for a research brief.

    EXAMPLE REQUEST
    Create a short brief from these notes. Source A reports that 18 of 30 participants preferred asynchronous feedback. Source B reports that interviews raised concerns about delayed clarification. Explain the tension and identify a question for follow-up.

    EXAMPLE RESULT
    Preference and responsiveness appear to pull in different directions. In Source A, 18 of 30 participants preferred asynchronous feedback. Source B's interviews describe concerns about waiting for clarification. Together, the supplied notes suggest that timing and communication preferences deserve separate attention. A useful follow-up question is whether a defined response window could preserve flexibility while reducing uncertainty. Sources: A; B. Open question: how did participants define an acceptable response time?

    MORE ABOUT THE WORKFLOW
    Transform your research. A clear, structured brief helps organize your research. ResearchBrief brings your research notes into a clear, structured format.

    LIMITATIONS AND HELP
    Check the result before using it. The page has a contact-the-creator button. No further support scope or update commitment is stated.

    PRICE AND PLATFORM INFORMATION
    The price panel shows US$12. The supplied platform documentation identifies this product's purchase control as a single payment; subscriptions use a separate recurring-plan selector. Its payment summary confirms a single US$12 charge, with applicable taxes calculated before payment and no automatic renewal.

    Accessible terms grant permanent personal use through the buyer's connected agent. The package is not supplied as a downloadable file. Refunds for a product not performing as described can be requested within 30 days under this fictional platform's terms. Any charges for the buyer's AI provider are separate. There are no extra paid services required by this offer. There are no tiers or quantity limits.

    NEXT ACTION
    Buy access.

    • Read your context and instructions
    • Compiled the gourmetskill

    Summary

    Listing diagnostic: 67.5/100. This is an editorial assessment of the fictional ResearchBrief listing supplied in the request, not a buyer rating or a test of the underlying skill.

    Coverage: complete supplied text for the main case; all eight dimensions observed, weighted coverage 100%. Confidence is high about what this supplied text contains, but that does not establish market predictive validity. No live website, package execution, or customer result was inspected.

    The useful deliverables are clear. The main gaps are the evidence for online source checking and the setup needed to perform it. The supplied purchase path makes the US$12 single payment and permanent personal hosted access understandable.

    Score breakdown

    DimensionWeightStatusAnchor / 4PointsEvidence and reason
    D1 — Reader and job fit10Observed37.5ABOUT identifies researchers, source notes and a brief. A useful non-use case or fit boundary is absent.
    D2 — Offer and deliverables20Observed315ABOUT identifies hosted access, a draft, outline, source list and open questions. Version and included/excluded capability boundaries are not developed.
    D3 — Claims and evidence20Observed210ABOUT claims online source checking. EXAMPLE RESULT demonstrates organization of supplied notes, not an online verification or unsupported-claim check. The claim is insufficiently demonstrated, not proven false.
    D4 — Requirements and setup15Observed27.5HOW TO START names hosts and a starting action. The browsing tools and source access needed for online checks are unspecified.
    D5 — Usage and output example15Observed311.25EXAMPLE REQUEST and EXAMPLE RESULT connect input to a concrete brief. A specific acceptance or failure check is absent.
    D6 — Price and terms clarity10Observed410PRICE AND PLATFORM INFORMATION supplies a single US$12 charge, no renewal, personal permanent hosted access, tax treatment, provider costs, refund conditions and no tiers. Repeating this in ABOUT is unnecessary.
    D7 — Limits and support5Observed22.5LIMITATIONS AND HELP supplies a general review reminder and contact route. Online-verification limitations, support scope and update expectations remain open.
    D8 — Structure and next action5Observed33.75Main facts and Buy access are findable. MORE ABOUT THE WORKFLOW repeats generic language instead of adding decision information.
    Total100All observed—67.5Sum of weight × anchor ÷ 4.

    Calculation: 7.5 + 15 + 10 + 7.5 + 11.25 + 10 + 2.5 + 3.75 = 67.5.

    Section comments

    SectionKeepImprove or verify
    Product name and summaryThe research-writing focus.Replace the broad transformation promise with a specific input and deliverable. D1/D8.
    AboutAudience, output inventory and hosted delivery.State a useful fit boundary; support the online-checking claim with a relevant example. D1/D3.
    How to startHost names and the paste-notes starting action.Confirm the actual tools and access required for online verification. D4.
    Example request and resultThe visible relationship between notes, source labels, synthesis and a follow-up question.Add an acceptance check. Do not present this example as proof of web verification. D5.
    More about the workflowNo unique information needs preserving.Remove repetition; use this space for scope or a relevant boundary. D8.
    Limitations and helpThe contact route and human-review reminder.Define support/update expectations and the fallback when sources are inaccessible. D7.
    Price and platform informationThe clear transaction and access conditions.No disclosure correction is needed on this supplied evidence. D6.
    Next actionBuy access accurately describes the hosted offer.Keep it. D8.

    Prioritized fixes

    PriorityLocationBuyer's unanswered questionActionOwner verification needed
    P1Online-checking claimWhat is checked, against which source, and what happens when access fails?Add a claim/source/result example with conditions and an unresolved-source case, or narrow the claim to what the current example shows.Actual online-checking behavior and limits.
    P1How to startWhich tools and access does online verification need?Name the real prerequisites and a minimal successful first run.Host tools, source access and any additional cost.
    P1Limitations and helpWhat help is included, and what should I do when verification fails?State the support scope and a specific fallback.Actual support/update policy and fallback behavior.
    P2Summary and repeated paragraphWhat do I get from my notes?Use the replacement copy below and remove the repeated slogan.No new performance claim is required.

    No P0 contradiction was established. Priorities describe disclosure and editorial risk, not expected conversion gains.

    Replacement copy

    Summary

    Turn supplied notes into a structured research brief, with claims linked to the sources you provide.

    Scope paragraph

    For researchers turning source notes and a question into a concise brief. Request a draft, section outline, source list and open questions. The example below demonstrates organization of supplied material; it does not demonstrate independent online verification.

    Example acceptance check

    Check that the brief retains the supplied findings, labels their sources, and distinguishes reported observations from a proposed follow-up question. In this example, a defined response window is a question to investigate, not an established result.

    These edits clarify what the supplied example supports. They do not prove online-checking performance or replace the need to confirm setup and support facts. No post-edit score or commercial uplift is claimed.

    Separate access-limit check

    If the entire purchase-information section is unavailable to the reviewer,

    D6 becomes Unknown, not zero or a low score. The other seven dimensions remain assessable on the supplied material, giving

    90% weighted coverage.

    Overall score: withheld. Request the relevant purchase panel or accessible terms. Missing reviewer access does not establish buyer confusion. Do not publish a normalized partial total as a whole-page score.

    Commercial outcomes

    PlatformReal buyer rating — first 30 / 90 daysPurchases — first 30 / 90 daysDownloads — first 30 / 90 days
    Fictional example marketplaceData insufficient / Data insufficientData insufficient / Data insufficientData insufficient / Data insufficient

    There is no validated outcome model. Hosted access is not a download event.

    Next steps: use the bounded replacement copy, confirm online-checking prerequisites and support scope, add the missing evidence, then request a fresh review of the revised complete listing.

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

    About this skill

    A discerning second look at your AI skill listing

    Give GourmetSkill your listing and price. Get a reasoned review, prioritized fixes, and replacement copy you can use.

    For skill creators preparing a launch or revising an existing offer. It helps a buyer understand who the skill is for, what they receive, how to start, and what the purchase means.

    What you receive

    Access to the GourmetSkill v1.1.0 listing-review workflow through your Agensi-connected agent, including its rubric, evidence-handling rules, and report structure.

    Each review can provide:

    • A listing diagnostic out of 100, when all applicable dimensions are assessable.
    • Section-by-section comments, with the passage behind each judgment.
    • Prioritized actions: P0 for contradictions, P1 for information needed to evaluate the offer, and P2 for readability.
    • Replacement wording for the most important passages, preserving supported facts.
    • Clear labels for observed information, owner-confirmed facts, and unresolved details.
    DimensionWeight
    Reader and job fit10
    Offer and deliverables20
    Claims and inspectable evidence20
    Requirements and setup15
    Usage and output example15
    Price and terms clarity10
    Limits and support expectations5
    Structure and next action5

    Each dimension uses a 0–4 anchor. Its contribution is weight × anchor ÷ 4. The weights are editorial choices, not a marketplace's official scoring formula.

    Fair to the information buyers actually see

    Platform price controls and accessible purchase terms count alongside the description. Clear information does not need to be repeated just to earn points.

    Not visible to the reviewer does not mean missing for the buyer. If a decisive detail cannot be inspected, GourmetSkill marks it unknown. If any applicable dimension is unknown, it withholds the overall score and reports assessable feedback and coverage.

    Your first review

    1. Connect your agent to Agensi and load GourmetSkill by name.
    2. Supply the listing URL or complete text, current price and billing model, and relevant buyer-facing terms. Mention any sections you have omitted.
    3. Ask for the diagnostic, section comments, and replacement copy in your preferred language.

    Success check: the report identifies its coverage, ties judgments to supplied passages, shows the score calculation when eligible, and separates supported edits from facts that still need verification.

    If a URL is blocked or a purchase panel cannot be read, paste the relevant text or supply a readable screenshot if your agent supports images. Do not share credentials. A pasted-text review requires no browsing, extra runtime, or skill-specific API key; your AI provider's own charges may still apply.

    Inspect the worked example below

    The demo reviews a fictional ResearchBrief listing supplied as complete text. Its 67.5/100 diagnosis is traceable to eight judgments, with replacement copy and a separate access-limit check. It demonstrates the review method; it is not a customer testimonial, a live browser test, or evidence of improved sales. Scores can vary with the agent and available evidence.

    Scope, help, and updates

    GourmetSkill reviews listing information and price presentation. It does not test the skill package, certify compatibility, determine an optimal price, or predict buyer ratings, purchases, or downloads. Human review is needed before publishing changed claims or terms.

    This is a self-guided agent workflow, not a human consulting service. Use the included rubric and report guidance to question a judgment. For connection or access problems, see Agensi troubleshooting. No future feature or support response-time commitment is part of this offer.

    US$6.89 — one-time purchase

    Agensi supplies a permanent personal licence to use the skill through your own agent; the purchase does not include a downloadable copy of the skill. Approved updates are available under Agensi's use and licence rules. Applicable taxes are added at checkout; AI-provider usage charges are separate. See payment details.

    If the skill does not perform as described when sold, Agensi's conditional 30-day money-back guarantee applies; it is not an unconditional satisfaction guarantee.

    Buy access, load GourmetSkill in your connected agent, and bring the listing you want to improve.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Prompt & Skill Engineering.

    Version 1.1.0. Reviews listing disclosure, not package execution. The rubric is editorial and not market calibrated. Unknown evidence is not a defect. Human review is required before publishing revised claims or terms. Examples are fictional; commercial predictions remain Data insufficient. Screenshots and video are edited illustrations of the fictional case, not live screen recordings.

    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 buying a skill.

    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 6 days ago

    Nothing to install. It runs in your AI agent.

    No skill-specific software, runtime, or API key. Paste complete listing text if browsing is unavailable. Your AI provider’s own plan or usage charges may apply.

    Permissions

    Read Files
    Write Files
    Browser
    Network Access

    Read bundled references and user-selected inputs. Browse supplied listing URLs and relevant platform terms when needed. Save reports only on request to the chosen path. No shell, environment variables, private account credentials, or execution of the reviewed skill is required.

    Listed7 days ago

    What's inside

    Frequently Asked Questions

    It measures how well a listing explains its offer under eight editorial dimensions. It is not a customer rating, an official Agensi score, or a test of the skill package.

    No. Version 1.1.0 reports Data insufficient for those outcomes at both 30 and 90 days. A higher diagnostic score does not imply a commercial uplift.

    Paste the relevant content, or provide a readable screenshot if your agent supports images. Unseen evidence is marked unknown. The overall score is withheld if an applicable dimension cannot be assessed.

    No. Clear platform controls and accessible purchase terms count as evidence. Optional repetition is a wording choice, not a scoring requirement.

    Connect your AI agent to Agensi through MCP, then ask it to use GourmetSkill on your task. Your agent loads the latest approved version each time. Paid skills require a purchase or active subscription. Skill files are not available for download; your licence does not allow you to save, copy or export any part of a skill.

    Yes, you can request a review and replacement copy in your preferred language, subject to your host agent’s capabilities. The skill instructions are written in English.

    The workflow includes its review rules and report guidance. For inaccessible page content, use the pasted-text fallback; for Agensi access issues, use Agensi troubleshooting. Approved updates follow Agensi’s platform policy. No future feature or support response time is promised.

    New in Prompt & Skill Engineering