Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Get an external patent examiner review of a patent application. Use when user says "专利审查", "patent review", "审查意见", "examiner review", or wants critical feedback on patent claims and specification.
.claude/skills/wanshuiyin-patent-review/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 51% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 46% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 77% | 0% |
Get a multi-round patent examiner review of the patent application based on: $ARGUMENTS
Adapted from /research-review. The reviewer persona is a patent examiner, not a paper reviewer.
REVIEWER_MODEL = gpt-6-astra — Model used via Codex MCPREVIEW_ROUNDS = 2 — Number of review roundsEXAMINER_PERSONA = "patent-examiner" — GPT-6-Astra personabash claude mcp add codex -s user -- python3 "$HOME/aris_repo/mcp-servers/codex-exec/server.py" # your ARIS clone's path
patent/CLAIMS.md — all drafted claimspatent/specification/ — all specification sectionspatent/figures/numeral_index.md — reference numeral mappingpatent/PRIOR_ART_REPORT.md — known prior artpatent/INVENTION_DISCLOSURE.md — invention structureBefore calling the external reviewer, compile a comprehensive briefing:
Send to REVIEWER_MODEL via spawn_agent with xhigh reasoning:
textspawn_agent: model: gpt-6-astra reasoning_effort: xhigh message: | You are a senior patent examiner at the [USPTO/CNIPA/EPO]. Examine this patent application and issue a detailed office action. CLAIMS: [all claims] SPECIFICATION SUMMARY: [key sections: title, technical field, background, summary, abstract] PRIOR ART KNOWN: [prior art references] PATENTABILITY STANDARDS TO APPLY: [US: 35 USC 101/102/103/112 | CN: Articles 22, 26 | EP: Articles 54, 56, 83, 84] Please issue an office action covering: 1. CLAIM CLARITY (112(b)/Art 84): - Are all terms definite? - Any indefinite functional language? - Antecedent basis issues? 2. WRITTEN DESCRIPTION (112(a)/Art 83 first para): - Does the spec support ALL claim scope? - Any claim elements without spec support? 3. ENABLEMENT (112(a)/Art 83): - Can a POSITA practice the invention? - Any missing algorithm/structure for functional claims? 4. NOVELTY (102/Art 54): - Would any known reference anticipate any claim? - Identify the closest single reference. 5. NON-OBVIOUSNESS (103/Art 56): - Would any combination render claims obvious? - What is the motivation to combine? 6. CLAIM SCOPE: - Are independent claims broad enough to be commercially valuable? - Do dependent claims provide meaningful fallback positions? - Any claims that are too broad (likely rejected) or too narrow (not valuable)? 7. SPECIFICATION QUALITY: - Language issues (subjective terms, relative terms, result-to-be-achieved) - Reference numeral consistency - Missing embodiments Format your response as a formal office action with: - GROUNDS OF REJECTION for each issue (cite statute) - SUGGESTED AMENDMENTS for each issue - OVERALL PATENTABILITY SCORE: 1-10 Be rigorous and specific. This is a real examination.
Based on the examiner's office action:
For each fix:
Use send_input with the saved reviewer id from Round 1:
textsend_input: target: [saved reviewer id from Round 1] message: | Here is the revised patent application after addressing your office action. CHANGES MADE: [list of all changes with rationale] REVISED CLAIMS: [updated claims] REVISED SPECIFICATION EXCERPTS: [changed sections] Please re-examine: 1. Are the previous rejections overcome? 2. Are there new issues introduced by the amendments? 3. What is the updated patentability score? 4. Any remaining grounds for rejection?
Write patent/PATENT_REVIEW.md:
markdown## Patent Review Report ### Application Summary [Title, claims count, jurisdiction] ### Review Round 1 #### Office Action Summary [Key findings from examiner] #### Issues Found | # | Type | Severity | Claim/Section | Issue | Citation | Fix Applied | |---|------|----------|--------------|-------|----------|-------------| | 1 | Clarity | CRITICAL | Claim 3 | Indefinite term "rapid" | 112(b) | Defined in spec | | 2 | Novelty | MAJOR | Claim 1 | Ref X anticipates element C | 102 | Amended claim | #### Score After Round 1: [X]/10 ### Review Round 2 #### Follow-Up Assessment [Are previous rejections overcome?] #### Remaining Issues [Any issues still outstanding] #### Score After Round 2: [X]/10 ### Recommendations [Final recommendations before proceeding to jurisdiction formatting] - [ ] All CRITICAL issues resolved - [ ] All MAJOR issues resolved or argued - [ ] Specification supports all claim amendments - [ ] Ready for jurisdiction formatting
model_reasoning_effort: "xhigh" for maximum analysis depth.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 16,598 | 16,436 | -1% | 1 | 1 | 0% | 262 | 1,897 | +624% | 0 | 0 | — |
case-02 | fail→fail | 15,644 | 50,939 | +226% | 1 | 1 | 0% | 214 | 2,071 | +868% | 0 | 0 | — |
case-03 | fail→fail | 16,921 | 13,168 | -22% | 1 | 1 | 0% | 294 | 1,997 | +579% | 0 | 0 | — |
case-04 | fail→fail | 23,182 | 18,377 | -21% | 1 | 1 | 0% | 3,065 | 1,869 | -39% | 0 | 0 | — |
case-05 | pass→pass | 36,246 | 55,571 | +53% | 1 | 1 | 0% | 2,193 | 6,739 | +207% | 0 | 0 | — |
case-06 | fail→fail | 5,716 | 19,584 | +243% | 1 | 1 | 0% | 410 | 2,298 | +460% | 0 | 0 | — |
case-07 | fail→pass | 15,931 | 9,350 | -41% | 1 | 1 | 0% | 1,928 | 2,267 | +18% | 0 | 0 | — |
case-08 | fail→pass | 19,144 | 12,701 | -34% | 1 | 1 | 0% | 1,895 | 2,858 | +51% | 0 | 0 | — |
case-09 | fail→pass | 16,941 | 12,884 | -24% | 1 | 1 | 0% | 2,072 | 2,636 | +27% | 0 | 0 | — |
case-10 | fail→pass | 17,704 | 11,368 | -36% | 1 | 1 | 0% | 1,729 | 2,530 | +46% | 0 | 0 | — |
case-11 | pass→pass | 17,845 | 7,374 | -59% | 1 | 1 | 0% | 2,054 | 1,917 | -7% | 0 | 0 | — |
case-17 | pass→pass | 36,397 | 27,116 | -25% | 1 | 1 | 0% | 2,589 | 4,348 | +68% | 0 | 0 | — |
case-12 | fail→pass | 12,308 | 19,445 | +58% | 1 | 1 | 0% | 2,424 | 4,291 | +77% | 0 | 0 | — |
case-13 | fail→pass | 20,718 | 16,622 | -20% | 1 | 1 | 0% | 2,170 | 3,424 | +58% | 0 | 0 | — |
case-14 | fail→pass | 11,826 | 8,406 | -29% | 1 | 1 | 0% | 850 | 2,191 | +158% | 0 | 0 | — |
case-15 | fail→pass | 6,257 | 13,412 | +114% | 1 | 1 | 0% | 998 | 2,749 | +175% | 0 | 0 | — |
case-16 | fail→fail | 9,886 | 12,007 | +21% | 1 | 1 | 0% | 1,450 | 2,114 | +46% | 0 | 0 | — |
case-18 | pass→pass | 19,301 | 16,422 | -15% | 1 | 1 | 0% | 2,412 | 3,542 | +47% | 0 | 0 | — |
case-19 | fail→pass | 17,748 | 10,696 | -40% | 1 | 1 | 0% | 2,282 | 2,425 | +6% | 0 | 0 | — |
case-20 | fail→fail | 19,358 | 13,714 | -29% | 1 | 1 | 0% | 2,067 | 2,906 | +41% | 0 | 0 | — |
case-21 | fail→pass | 12,179 | 13,615 | +12% | 1 | 1 | 0% | 1,931 | 2,830 | +47% | 0 | 0 | — |
case-22 | fail→pass | 13,043 | 8,803 | -33% | 1 | 1 | 0% | 1,119 | 2,212 | +98% | 0 | 0 | — |
case-23 | fail→pass | 10,824 | 9,827 | -9% | 1 | 1 | 0% | 1,467 | 2,381 | +62% | 0 | 0 | — |
case-24 | fail→pass | 11,537 | 9,692 | -16% | 1 | 1 | 0% | 1,788 | 3,151 | +76% | 0 | 0 | — |
case-25 | fail→pass | 11,551 | 6,019 | -48% | 1 | 1 | 0% | 1,735 | 2,702 | +56% | 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. 25 cases were attempted, and 20 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 +56 percentage points is the difference between those two pass rates over the 20 comparable cases. 2 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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 9/9/2026 | +22% |
| gemini-3.6-flash | verified | 8/11/2026 | +30% |
Other measured skills in the registry, with their headline benchmark lift.