Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Analyze competitor IP portfolios — comparative patent portfolio reports with strategy inference. Budget: 120 patent families, 15 claim parses, 40 web searches.
.claude/skills/yogsoth-ai-competitive-intelligence/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 136% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 161% | 0% |
Analyzes competitor patent portfolios to infer R&D strategy, identify strengths and weaknesses, and produce comparative intelligence reports.
Profile competitor IP portfolios to understand their technology focus, filing strategy, geographic coverage, and potential future directions. Enables strategic IP positioning.
| Metric | Target | |--------|--------| | Patent families analyzed | 120 | | Claim parses | 15 | | Web searches completed | 40 |
| Metric | Target | Current | % | |--------|--------|---------|---| | Patent families analyzed | 120 | 0 | 0% | | Claim parses completed | 15 | 0 | 0% | | Web searches completed | 40 | 0 | 0% | | Competitors profiled | — | 0 | — | | Portfolio comparisons | — | 0 | — |
HARD-GATE: Cannot exit iteration loop until 80% of patent families (96), claim parses (12), and web searches (32) budget met.
| Tactic | When to Use | |--------|-------------| | patent-family-tracing | Build complete portfolio for each competitor | | classification-navigation | Map competitor technology focus areas | | claim-decomposition | Analyze claim scope of competitor key patents |
| SOP | Role in This Strategy | |-----|----------------------| | patent-query-formulation | Generate assignee-focused search queries | | assignee-normalization | Resolve subsidiary/parent relationships | | patent-categorization | Classify competitor patents by technology area | | citation-network-analysis | Map inter-competitor citation relationships | | trend-analysis | Analyze competitor filing velocity and direction | | claim-parsing | Parse key competitor claims for scope analysis | | quality-scoring | Assess competitor patent quality distribution | | legal-status-assessment | Determine active vs. expired competitor IP | | saturation-detection | Confirm portfolio coverage is complete | | patent-synthesis | Produce competitive intelligence report |
assignee-normalizationpatent-query-formulationpatent-family-tracingpatent-categorizationtrend-analysiscitation-network-analysisclaim-parsingquality-scoringpatent-synthesismarkdown# Competitive Patent Intelligence: [Technology Domain] ## Competitor Profiles ### [Competitor A] - Portfolio size: [N] families - Technology focus: [top IPC classes] - Filing trend: [increasing/stable/declining] - Geographic coverage: [jurisdictions] - Key patents: [top 3 by quality score] ## Comparative Analysis | Metric | Comp A | Comp B | Comp C | |--------|--------|--------|--------| | Total families | | | | | Filing velocity (last 3yr) | | | | | Avg quality score | | | | | Geographic breadth | | | | ## Citation Network [Inter-competitor citation patterns and influence] ## Strategic Inferences [R&D direction, potential acquisitions, licensing opportunities] ## Recommendations [Positioning strategy relative to competitors]
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | classification-navigation | IPC/CPC hierarchy drill-down and lateral expansion for patent discovery | | knowledge-acquisition-claim-decomposition | Independent/dependent claim parsing, element extraction, and feature mapping to technical domains | | patent-family-tracing | Forward/backward patent citation and priority tracing until saturation |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | assignee-normalization | Standardize assignee names and identify corporate group affiliations across patent offices | | citation-network-analysis | Build and analyze patent citation networks — main path analysis, PageRank, cluster detection | | claim-parsing | Patent claim syntax parsing — independent/dependent relationships and element extraction | | knowledge-acquisition-saturation-detection | Determine when additional searching yields diminishing returns. Analyzes the latest expansion batch against existing corpus to judge continue/near-saturation/saturated. Used by snowball and systematic-survey. | | legal-status-assessment | Determine patent legal status — active, expired, pending, lapsed, or revoked | | patent-categorization | Classify patents by tech subdomain, application scenario, and value chain position | | patent-query-formulation | Construct keyword + IPC/CPC + assignee combination search strategies for patent databases | | patent-synthesis | Produce final structured patent intelligence report from all analysis results | | quality-scoring | Multi-dimensional patent quality assessment — forward citations, family size, claim count, geographic breadth | | trend-analysis | Patent filing volume time-series, technology lifecycle stage, and S-curve analysis |
<!-- END available-tables (generated) -->
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 36,821 | 38,477 | +4% | 1 | 1 | 0% | 6,280 | 7,540 | +20% | 0 | 0 | — |
case-02 | fail→fail | 35,508 | 8,218 | -77% | 1 | 1 | 0% | 6,263 | 2,809 | -55% | 0 | 0 | — |
case-03 | fail→fail | 36,541 | 35,375 | -3% | 1 | 1 | 0% | 6,269 | 7,528 | +20% | 0 | 0 | — |
case-04 | pass→fail | 10,881 | 33,828 | +211% | 1 | 1 | 0% | 1,809 | 7,444 | +311% | 0 | 0 | — |
case-05 | pass→pass | 13,271 | 34,331 | +159% | 1 | 1 | 0% | 1,898 | 7,450 | +293% | 0 | 0 | — |
case-06 | pass→pass | 23,129 | 36,179 | +56% | 1 | 1 | 0% | 3,838 | 7,459 | +94% | 0 | 0 | — |
case-07 | fail→pass | 12,475 | 5,346 | -57% | 1 | 1 | 0% | 1,762 | 2,192 | +24% | 0 | 0 | — |
case-08 | fail→pass | 13,439 | 23,979 | +78% | 1 | 1 | 0% | 2,005 | 4,730 | +136% | 0 | 0 | — |
case-09 | pass→pass | 18,040 | 25,859 | +43% | 1 | 1 | 0% | 2,378 | 5,114 | +115% | 0 | 0 | — |
case-10 | pass→pass | 14,321 | 18,407 | +29% | 1 | 1 | 0% | 1,913 | 4,142 | +117% | 0 | 0 | — |
case-11 | fail→pass | 8,893 | 2,752 | -69% | 1 | 1 | 0% | 1,169 | 1,673 | +43% | 0 | 0 | — |
case-12 | pass→pass | 11,486 | 14,524 | +26% | 1 | 1 | 0% | 1,642 | 3,308 | +101% | 0 | 0 | — |
case-13 | pass→pass | 15,502 | 35,145 | +127% | 1 | 1 | 0% | 2,170 | 6,821 | +214% | 0 | 0 | — |
case-14 | pass→pass | 14,557 | 15,321 | +5% | 1 | 1 | 0% | 2,014 | 3,498 | +74% | 0 | 0 | — |
case-15 | fail→pass | 14,271 | 13,175 | -8% | 1 | 1 | 0% | 2,244 | 3,205 | +43% | 0 | 0 | — |
case-16 | fail→pass | 29,137 | 11,087 | -62% | 1 | 1 | 0% | 1,131 | 2,950 | +161% | 0 | 0 | — |
case-17 | pass→fail | 15,413 | 12,782 | -17% | 1 | 1 | 0% | 2,251 | 1,958 | -13% | 0 | 0 | — |
case-18 | pass→pass | 13,620 | 10,883 | -20% | 1 | 1 | 0% | 2,164 | 2,995 | +38% | 0 | 0 | — |
case-19 | pass→pass | 13,107 | 10,050 | -23% | 1 | 1 | 0% | 1,858 | 2,776 | +49% | 0 | 0 | — |
case-20 | pass→pass | 12,531 | 6,465 | -48% | 1 | 1 | 0% | 1,902 | 2,231 | +17% | 0 | 0 | — |
case-21 | pass→pass | 11,373 | 5,730 | -50% | 1 | 1 | 0% | 1,672 | 2,083 | +25% | 0 | 0 | — |
case-22 | fail→pass | 11,565 | 8,532 | -26% | 1 | 1 | 0% | 1,768 | 2,560 | +45% | 0 | 0 | — |
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 20 counted toward the lift figure. The other 2 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 +18 percentage points is the difference between those two pass rates over the 20 comparable cases. 4 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.