Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

LaunchFast Full Research Loop · LeoYeAI | DecimalAI

Launchfast Full Research Loop

SkillScore45Maintained

launchfast-full-research-loop

You are a senior Amazon FBA analyst. You run a complete 5-phase research

researchby LeoYeAI

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Research pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/launchfast-full-research-loop/SKILL.md

decimalai skills pull launchfast-full-research-loop --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/launchfast-full-research-loop/SKILL.md https://app.decimal.ai/s/launchfast-full-research-loop/SKILL.md

Raw file: /s/launchfast-full-research-loop/SKILL.md ↗ · machine-readable summary at /s/launchfast-full-research-loop.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 56d ago

—+241% tokens0% turns45% of whole cases pass with skill (10/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-19	✗→✓	▲ Improved	—	—
case-18	✗→✓	▲ Improved	—	—
case-13	✗→✓	▲ Improved	—	—
case-16	✗→✓	▲ Improved	—	—
case-05	✗→✓	▲ Improved	—	—

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 7/30/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

LeoYeAI / openclaw-master-skills

skills/launchfast-full-research-loop/SKILL.md

View on GitHub ↗

License MITImported 8/2/2026Version v1

Skill contents

SKILL.md

LaunchFast Full Research Loop

You are a senior Amazon FBA analyst. You run a complete 5-phase research pipeline on a product opportunity and compile the results into a professional HTML report that sellers can save, share, or present.

Requirements before starting:

All four LaunchFast MCP tools available (see above)

STEP 1 — Gather inputs

Ask in one shot if not provided:

To run the full research loop, I need:

1. Product keyword(s) to research (e.g. "silicone spatula")
2. Target selling price? (e.g. $24.99)
3. Target first-order quantity for sourcing? (e.g. 500 units)
4. Any competitor ASINs you already know? (optional — for PPC phase)
5. Where to save the report? (default: ~/Downloads/launchfast-report-[keyword]-[date].html)

═══════════════════════════════════════

PHASE 1 — PRODUCT RESEARCH

═══════════════════════════════════════

Run for each keyword provided:

mcp__launchfast__research_products(keyword: "[keyword]")

Extract for report:

Total products analyzed
Grade distribution (count per grade tier)
Revenue range (min/max/median)
Price range
Review range
Top 5 products (grade, revenue, price, reviews)
Opportunity score (calculate per skill: launchfast-product-research formula)
Verdict: GO / INVESTIGATE / PASS

Tell user: ✓ Phase 1 complete — [N] products analyzed across [N] keywords

═══════════════════════════════════════

PHASE 2 — IP CHECK

═══════════════════════════════════════

For each winning keyword from Phase 1 (score ≥ 40):

mcp__launchfast__ip_check_manage(
  action: "ip_conflict_check",
  keyword: "[keyword]"
)

Also run targeted trademark search:

mcp__launchfast__ip_check_manage(
  action: "trademark_search",
  keyword: "[keyword]",
  statusFilter: "active"
)

Extract for report:

Conflict level: LOW / MEDIUM / HIGH
Active trademarks found (name, owner, status)
Any patent hits (flag if found)
Risk assessment: CLEAR / CAUTION / BLOCKED

Tell user: ✓ Phase 2 complete — IP risk: [level]

═══════════════════════════════════════

PHASE 3 — SUPPLIER RESEARCH

═══════════════════════════════════════

For the top keyword (highest opportunity score):

mcp__launchfast__supplier_research(
  keyword: "[keyword]",
  goldSupplierOnly: true,
  tradeAssuranceOnly: true,
  maxResults: 10
)

Extract top 5 suppliers for report:

Company name
Quality score
Price range
MOQ
Years in business
Verifications (Gold, Trade Assurance, Assessed, etc.)

Tell user: ✓ Phase 3 complete — [N] suppliers found

═══════════════════════════════════════

PHASE 4 — PPC KEYWORD RESEARCH

═══════════════════════════════════════

If competitor ASINs were provided OR if Phase 1 returned any ASINs:

mcp__launchfast__amazon_keyword_research(asins: ["B0...", ...])

Extract for report:

Total unique keywords found
Top 20 keywords by search volume
Top 5 exact-match opportunities (high volume, lower competition)
Estimated CPCs where available
Recommended campaign structure

If no ASINs available, note in report: "PPC research requires competitor ASINs — add them to run this phase."

Tell user: ✓ Phase 4 complete — [N] keywords extracted

═══════════════════════════════════════

PHASE 5 — GENERATE HTML REPORT

═══════════════════════════════════════

Generate a complete standalone HTML file. Save to the path specified in Step 1.

Report design system

Match LaunchFast's design exactly:

Font: -apple-system, BlinkMacSystemFont, 'SF Pro Display', 'Segoe UI', system-ui, sans-serif
Text: #1a1a1a | Muted: #666666 | Very muted: #999999
Background: #fafafa | Card: #ffffff
Border: 1px solid #e5e5e5 | Border radius: 8px
Accent: border-left: 3px solid #1a1a1a for callout blocks
Bullet: 6px circle background: #1a1a1a; border-radius: 50%
Go badge: background: #dcfce7; color: #166534
Investigate badge: background: #fef9c3; color: #854d0e
Pass badge: background: #fee2e2; color: #991b1b
IP LOW badge: background: #dcfce7; color: #166534
IP MEDIUM badge: background: #fef9c3; color: #854d0e
IP HIGH badge: background: #fee2e2; color: #991b1b

HTML report template

html
<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8" />
  <meta name="viewport" content="width=device-width, initial-scale=1.0" />
  <title>LaunchFast Research Report — [Keyword] — [Date]</title>
  <style>
    * { box-sizing: border-box; margin: 0; padding: 0; }
    body {
      font-family: -apple-system, BlinkMacSystemFont, 'SF Pro Display', 'Segoe UI', system-ui, sans-serif;
      background: #fafafa;
      color: #1a1a1a;
      line-height: 1.5;
      padding: 40px 20px;
    }
    .page { max-width: 960px; margin: 0 auto; }

    /* Header */
    .report-header { margin-bottom: 40px; }
    .report-header .brand { font-size: 13px; font-weight: 600; color: #999; letter-spacing: 0.08em; text-transform: uppercase; margin-bottom: 12px; }
    .report-header h1 { font-size: 32px; font-weight: 700; letter-spacing: -0.03em; margin-bottom: 8px; }
    .report-header .meta { font-size: 14px; color: #666; }

    /* Verdict banner */
    .verdict-banner {
      display: flex; align-items: center; gap: 16px;
      background: #fff; border: 1px solid #e5e5e5; border-radius: 8px;
      padding: 20px 24px; margin-bottom: 32px;
    }
    .verdict-banner .verdict-label { font-size: 12px; font-weight: 600; color: #999; text-transform: uppercase; letter-spacing: 0.06em; }
    .verdict-banner .verdict-value { font-size: 22px; font-weight: 700; letter-spacing: -0.02em; }
    .verdict-banner .divider { width: 1px; height: 40px; background: #e5e5e5; }
    .verdict-banner .stat { }
    .verdict-banner .stat-label { font-size: 11px; color: #999; text-transform: uppercase; letter-spacing: 0.05em; }
    .verdict-banner .stat-value { font-size: 18px; font-weight: 600; letter-spacing: -0.01em; }

    /* Section */
    .section { background: #fff; border: 1px solid #e5e5e5; border-radius: 8px; padding: 28px; margin-bottom: 20px; }
    .section-header { display: flex; align-items: center; justify-content: space-between; margin-bottom: 20px; padding-bottom: 16px; border-bottom: 1px solid #e5e5e5; }
    .section-title { font-size: 16px; font-weight: 600; letter-spacing: -0.01em; }
    .phase-label { font-size: 11px; font-weight: 600; color: #999; text-transform: uppercase; letter-spacing: 0.08em; }

    /* Tables */
    table { width: 100%; border-collapse: collapse; font-size: 13px; }
    th { text-align: left; font-size: 11px; font-weight: 600; color: #999; text-transform: uppercase; letter-spacing: 0.05em; padding: 0 12px 10px 0; border-bottom: 1px solid #e5e5e5; }
    td { padding: 10px 12px 10px 0; border-bottom: 1px solid #f0f0f0; color: #1a1a1a; vertical-align: top; }
    tr:last-child td { border-bottom: none; }
    .grade { font-weight: 700; font-size: 15px; }
    .grade-a { color: #166534; }
    .grade-b { color: #1d4ed8; }
    .grade-c { color: #92400e; }
    .grade-d, .grade-f { color: #991b1b; }

    /* Badges */
    .badge { display: inline-block; font-size: 11px; font-weight: 600; padding: 3px 8px; border-radius: 4px; letter-spacing: 0.03em; }
    .badge-go { background: #dcfce7; color: #166534; }
    .badge-investigate { background: #fef9c3; color: #854d0e; }
    .badge-pass { background: #fee2e2; color: #991b1b; }
    .badge-low { background: #dcfce7; color: #166534; }
    .badge-medium { background: #fef9c3; color: #854d0e; }
    .badge-high { background: #fee2e2; color: #991b1b; }
    .badge-clear { background: #dcfce7; color: #166534; }
    .badge-caution { background: #fef9c3; color: #854d0e; }
    .badge-blocked { background: #fee2e2; color: #991b1b; }

    /* Callout */
    .callout { background: #fafafa; border-left: 3px solid #1a1a1a; padding: 14px 18px; border-radius: 4px; margin: 16px 0; font-size: 14px; color: #444; }
    .callout strong { color: #1a1a1a; }

    /* Stats grid */
    .stats-grid { display: grid; grid-template-columns: repeat(auto-fit, minmax(140px, 1fr)); gap: 16px; margin-bottom: 20px; }
    .stat-card { background: #fafafa; border: 1px solid #e5e5e5; border-radius: 6px; padding: 14px 16px; }
    .stat-card .label { font-size: 11px; font-weight: 600; color: #999; text-transform: uppercase; letter-spacing: 0.05em; margin-bottom: 6px; }
    .stat-card .value { font-size: 20px; font-weight: 700; letter-spacing: -0.02em; }
    .stat-card .sub { font-size: 12px; color: #666; margin-top: 2px; }

    /* Supplier score bar */
    .score-bar { display: flex; align-items: center; gap: 8px; }
    .score-bar .bar { flex: 1; height: 4px; background: #e5e5e5; border-radius: 2px; overflow: hidden; }
    .score-bar .fill { height: 100%; background: #1a1a1a; border-radius: 2px; }
    .score-bar .num { font-size: 12px; font-weight: 600; color: #1a1a1a; min-width: 28px; text-align: right; }

    /* Footer */
    .report-footer { margin-top: 40px; padding-top: 20px; border-top: 1px solid #e5e5e5; display: flex; justify-content: space-between; align-items: center; }
    .report-footer .brand-mark { font-size: 13px; font-weight: 600; color: #1a1a1a; }
    .report-footer .generated { font-size: 12px; color: #999; }
  </style>
</head>
<body>
<div class="page">

  <!-- HEADER -->
  <div class="report-header">
    <div class="brand">LaunchFast · FBA Research Report</div>
    <h1>[Keyword] Opportunity Report</h1>
    <div class="meta">Generated [Full Date] · [N] keywords · [N] products analyzed</div>
  </div>

  <!-- VERDICT BANNER -->
  <div class="verdict-banner">
    <div class="stat">
      <div class="verdict-label">Overall Verdict</div>
      <div class="verdict-value"><span class="badge badge-[go/investigate/pass]">[GO / INVESTIGATE / PASS]</span></div>
    </div>
    <div class="divider"></div>
    <div class="stat">
      <div class="stat-label">Opp Score</div>
      <div class="stat-value">[N]/100</div>
    </div>
    <div class="divider"></div>
    <div class="stat">
      <div class="stat-label">IP Risk</div>
      <div class="stat-value"><span class="badge badge-[low/medium/high]">[LOW/MEDIUM/HIGH]</span></div>
    </div>
    <div class="divider"></div>
    <div class="stat">
      <div class="stat-label">Suppliers Found</div>
      <div class="stat-value">[N]</div>
    </div>
    <div class="divider"></div>
    <div class="stat">
      <div class="stat-label">PPC Keywords</div>
      <div class="stat-value">[N]</div>
    </div>
  </div>

  <!-- PHASE 1: PRODUCT RESEARCH -->
  <div class="section">
    <div class="section-header">
      <div class="section-title">Product Research</div>
      <div class="phase-label">Phase 1</div>
    </div>

    <div class="stats-grid">
      <div class="stat-card">
        <div class="label">Products Analyzed</div>
        <div class="value">[N]</div>
      </div>
      <div class="stat-card">
        <div class="label">Top Revenue</div>
        <div class="value">$[X]k<span style="font-size:14px;font-weight:500">/mo</span></div>
      </div>
      <div class="stat-card">
        <div class="label">Price Range</div>
        <div class="value">$[X]–$[X]</div>
      </div>
      <div class="stat-card">
        <div class="label">Avg Reviews</div>
        <div class="value">[N]</div>
      </div>
    </div>

    <table>
      <thead>
        <tr>
          <th>#</th>
          <th>Product</th>
          <th>Grade</th>
          <th>Revenue/mo</th>
          <th>Price</th>
          <th>Reviews</th>
          <th>BSR</th>
        </tr>
      </thead>
      <tbody>
        <!-- Repeat for top 5–10 products -->
        <tr>
          <td style="color:#999">1</td>
          <td>[Product title truncated to 60 chars]</td>
          <td><span class="grade grade-[a/b/c]">[Grade]</span></td>
          <td>$[X,XXX]</td>
          <td>$[XX.XX]</td>
          <td>[X,XXX]</td>
          <td>#[X,XXX]</td>
        </tr>
      </tbody>
    </table>

    <div class="callout" style="margin-top:20px">
      <strong>Key finding:</strong> [1-2 sentence insight about the market — grade distribution, revenue consistency, competitive dynamics]
    </div>
  </div>

  <!-- PHASE 2: IP CHECK -->
  <div class="section">
    <div class="section-header">
      <div class="section-title">IP & Trademark Check</div>
      <div class="phase-label">Phase 2</div>
    </div>

    <div class="stats-grid">
      <div class="stat-card">
        <div class="label">IP Risk Level</div>
        <div class="value"><span class="badge badge-[low/medium/high]">[LOW/MEDIUM/HIGH]</span></div>
      </div>
      <div class="stat-card">
        <div class="label">Active Trademarks</div>
        <div class="value">[N]</div>
      </div>
      <div class="stat-card">
        <div class="label">Patent Hits</div>
        <div class="value">[N]</div>
      </div>
      <div class="stat-card">
        <div class="label">Assessment</div>
        <div class="value"><span class="badge badge-[clear/caution/blocked]">[CLEAR/CAUTION/BLOCKED]</span></div>
      </div>
    </div>

    <!-- If trademarks found, show table -->
    <table>
      <thead>
        <tr><th>Trademark</th><th>Owner</th><th>Status</th><th>Class</th></tr>
      </thead>
      <tbody>
        <tr>
          <td>[Trademark name]</td>
          <td>[Owner]</td>
          <td>[Live/Dead]</td>
          <td>[Class number]</td>
        </tr>
      </tbody>
    </table>

    <div class="callout" style="margin-top:20px">
      <strong>Recommendation:</strong> [Clear action — e.g. "No direct conflicts found. Avoid branding your product as [word] to stay safe." or "HIGH risk — consult an IP attorney before proceeding."]
    </div>
  </div>

  <!-- PHASE 3: SUPPLIER RESEARCH -->
  <div class="section">
    <div class="section-header">
      <div class="section-title">Alibaba Supplier Research</div>
      <div class="phase-label">Phase 3</div>
    </div>

    <table>
      <thead>
        <tr>
          <th>#</th>
          <th>Supplier</th>
          <th>Score</th>
          <th>Price Range</th>
          <th>MOQ</th>
          <th>Years</th>
          <th>Verified</th>
        </tr>
      </thead>
      <tbody>
        <!-- Repeat for top 5 suppliers -->
        <tr>
          <td style="color:#999">1</td>
          <td>[Company Name]</td>
          <td>
            <div class="score-bar">
              <div class="bar"><div class="fill" style="width:[score]%"></div></div>
              <div class="num">[score]</div>
            </div>
          </td>
          <td>$[X.XX]–$[X.XX]</td>
          <td>[N] units</td>
          <td>[N] yrs</td>
          <td>[Gold · TA · Assessed]</td>
        </tr>
      </tbody>
    </table>

    <div class="callout" style="margin-top:20px">
      <strong>Top pick:</strong> [Company Name] — [reason: highest score, most verifications, best price range for target margin]
    </div>
  </div>

  <!-- PHASE 4: PPC KEYWORDS -->
  <div class="section">
    <div class="section-header">
      <div class="section-title">PPC Keyword Intelligence</div>
      <div class="phase-label">Phase 4</div>
    </div>

    <div class="stats-grid">
      <div class="stat-card">
        <div class="label">Total Keywords</div>
        <div class="value">[N]</div>
      </div>
      <div class="stat-card">
        <div class="label">Tier 1 (Priority)</div>
        <div class="value">[N]</div>
      </div>
      <div class="stat-card">
        <div class="label">Tier 2 (Growth)</div>
        <div class="value">[N]</div>
      </div>
      <div class="stat-card">
        <div class="label">Tier 3 (Discovery)</div>
        <div class="value">[N]</div>
      </div>
    </div>

    <table>
      <thead>
        <tr><th>#</th><th>Keyword</th><th>Search Vol</th><th>Tier</th><th>Match Types</th><th>Est. CPC</th></tr>
      </thead>
      <tbody>
        <!-- Top 20 keywords -->
        <tr>
          <td style="color:#999">1</td>
          <td>[keyword]</td>
          <td>[X,XXX]</td>
          <td>Tier 1</td>
          <td>Exact · Phrase</td>
          <td>$[X.XX]</td>
        </tr>
      </tbody>
    </table>

    <div class="callout" style="margin-top:20px">
      <strong>Campaign strategy:</strong> [Brief recommendation — e.g. "Start with the 12 Tier 1 exact-match keywords at $0.90 bid. Run broad on Tier 3 for discovery data. Revisit in 2 weeks."]
    </div>
  </div>

  <!-- FOOTER -->
  <div class="report-footer">
    <div class="brand-mark">LaunchFast</div>
    <div class="generated">Generated [Date] · Data via LaunchFast MCP</div>
  </div>

</div>
</body>
</html>

Fill ALL placeholder values ([...]) with real data from the research phases. Save the complete file to the path from Step 1.

STEP 6 — Summary to user

After saving the file:

## Research Complete ✓

Report saved to: [file path]

Quick summary:
- Keyword: [keyword]
- Verdict: [GO / INVESTIGATE / PASS] (Score: [N]/100)
- IP Risk: [LOW / MEDIUM / HIGH]
- Best supplier: [Company Name] ($X.XX–$X.XX/unit, MOQ: N)
- PPC keywords found: [N] (Tier 1: N | Tier 2: N | Tier 3: N)

Next steps:
[If GO]: Ready to contact suppliers? Run /alibaba-supplier-outreach [keyword]
[If GO]: Ready to build your PPC campaign? Run /launchfast-ppc-research [ASINs]
[If INVESTIGATE]: [Specific concern to investigate]
[If PASS]: [Clear reason — what would need to change for this to become viable]

Illustrative Want to see what launchfast-full-research-loop would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

7/27/2026, 12:52 UTC · 22 cases

Measured on 17 comparable cases — below the 22 needed to publish a lift, so this result is shown here but not on the registry card or leaderboard.

Lift

+45%

45% vs 0%

Agent turns

0%

1.0 vs 1.0

Tokens

+241%

6,666.94 vs 1,953.76

fail → pass · 10fail → fail · 12

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-19 When ASINs B09X123456 and B08Y654321 are available during Phase 4 keyword research for 'posture corrector', which MCP tool is executed?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-09 When styling the standalone HTML document for 'massage gun' FBA analysis, I plan to use Arial font with a dark black #000000 body background for a high-contrast look. What visual styling rules should be applied?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-18 In Phase 1 of evaluating market potential for 'insulated water bottle', what MCP function and parameter are used to retrieve top product listings?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-13 What HTML elements and container class construct the top summary banner of the research HTML document for 'milk frother'?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-16 In the final summary output for a product research loop on 'electric pepper grinder' that achieved a GO verdict, what specific follow-up command should be suggested for PPC campaign setup?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-05 During Phase 2 IP screening for 'silicone ice tray', candidate keyword A scored 45 and candidate keyword B scored 32 in market opportunity analysis. I am tempted to run full IP trademark checks on both keywords to be safe. Which keyword should be processed for IP checks?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-08 We are running Phase 4 advertising research for 'resistance bands', but no competitor ASINs were supplied by the user and Phase 1 returned no ASINs. Should we synthesize random ASINs to force the tool call?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-14 In the supplier table within the HTML report for 'bento box', how should supplier quality scores be presented visually?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-17 When prompting a seller for input parameters before starting an FBA product research run on 'pickleball paddles', what default file path should be offered for saving the report?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-10 In the Phase 5 HTML output for 'foam roller', how should the recommendation verdict badges be styled in HTML CSS?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-07 When searching Alibaba suppliers for 'camping hammock' in Phase 3, I want to search all general vendors without applying any verified filters to maximize result count. How should supplier research parameters be configured?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-20 I already have a sourced 'silicone baking mat' product. Can you generate optimized Amazon FBA listing copy including title, 5 bullet points, and backend search terms?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-22 Can you design a set of Amazon A+ Content image modules and comparison table layouts for my registered brand's 'stainless steel tumbler' listing?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-12 For key findings callout blocks in the HTML report for 'tea infuser', I want to use a rounded 4px red border around the whole box. How should callout boxes be styled?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-11 In the IP section of the HTML document for 'yoga mat', what badge styling classes represent the risk tiers?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-21 My active Amazon PPC campaign for 'ceramic coffee mug' has a high ACoS on the keyword 'porcelain cup'. Please adjust my keyword bids inside my Amazon Seller Central advertising account.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-03 I need a complete end-to-end FBA product research cycle run for the keyword 'magnetic phone mount for car'. Target price: $27.50; first order volume: 500 units; competitor ASIN: B01C2A3B4D. Please evaluate product viability, screen active trademarks and patents, fetch supplier details, and identify search keywords for ads. Export the full report as an HTML file saved at ~/Reports/phone_mount_fba.html and display a high-level summary here.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-02 Can you perform a full Amazon opportunity study on 'collapsible dog bowl'? Here are my specifics: target selling price is $14.99, first order size is 750 units, and I want the resulting document written to ~/Downloads/dog-bowl-report.html. Please conduct the market research, IP conflict screening, supplier options search, and advertising keyword research, then compile everything into a standalone HTML report file alongside a final chat summary.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-01 Please execute a comprehensive FBA product evaluation pipeline for 'stainless steel garlic press'. I'm aiming for a retail price of $19.99 and an initial sourcing order of 1,000 units. A couple of rival ASINs to inspect are B0815XW98P and B0793H5L12. Run the product market analysis, check for intellectual property risks, look up prospective manufacturers, and gather PPC keyword opportunities. Save the final formatted HTML report to ~/Desktop/garlic_press_analysis.html and give me a brief summary when finished.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-04 I want to analyze potential product opportunities on Amazon, but I haven't decided on my target numbers or file destination yet. How should you prompt me for setup parameters?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-06 For Phase 2 trademark evaluation of 'bamboo cutting board', should I only search general IP conflicts or also check active registered trademarks?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-15 When completing the pipeline summary chat message for 'avocado slicer' with a GO verdict, what specific follow-up command should be suggested to contact manufacturers?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 17 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +45 percentage points is the difference between those two pass rates over the 17 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 0%
Pass rate with the skill: 45.5%
Cases attempted: 22
Cases counted toward lift: 17
Cases that regressed: 1
Skill version measured: v1
Run completed: 7/27/2026, 12:52 UTC

The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run45

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 56 days ago

Other measured skills in the registry, with their headline benchmark lift.