borghei/Claude-Skills / research/patent
patent
VerifiedPatent research and IP landscape: prior-art search, claim mapping, landscape analysis, and freedom-to-operate. Use when conducting a prior-art search, mapping an IP landscape, evaluating patentability, or running an FTO analysis.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 8.4 KB |
| references/claim-mapping-and-landscape.md | markdown | 6.8 KB |
| references/freedom-to-operate-and-patentability.md | markdown | 8.0 KB |
| references/prior-art-search-strategy.md | markdown | 6.2 KB |
| scripts/claim_landscape_mapper.py | script | 7.4 KB |
| scripts/patentability_scorer.py | script | 10.1 KB |
| scripts/prior_art_search_planner.py | script | 7.0 KB |
references/claim-mapping-and-landscape.md
View on GitHub# Claim Mapping & Landscape Reference
Reference for analyzing patent claims and mapping IP landscapes.
## 1. Anatomy of a patent
- **Title** — short descriptor
- **Abstract** — public-facing summary (cannot be used to interpret claims)
- **Field of the invention** — broad area
- **Background** — prior art context
- **Summary** — high-level invention description
- **Detailed description** — full implementation (enablement requirement)
- **Drawings + figures**
- **Claims** — THE legal scope; what's actually protected
- **Bibliography** — cited prior art
The claims are what matter. Everything else supports them.
## 2. Claim types
### Independent claims
- Stand alone; don't reference other claims
- Define the broadest scope
- Usually 1-3 per patent
- Highest litigation value
### Dependent claims
- Reference and narrow an independent claim
- Add features ("the system of claim 1, wherein...")
- Provide fallback if independent claim is invalidated
### Method claims
- "A method for X comprising: step a, step b, step c"
- Process protection
- Often used for software/algorithms
### System claims
- "A system comprising: component 1, component 2"
- Physical/structural protection
### Apparatus claims
- "An apparatus for X comprising: component A, component B"
- Similar to system; mechanical bias
### Composition claims (chemistry)
- "A composition comprising: component A, B, C"
- Chemical formulations
### Use claims
- "Use of compound X for Y" (varies by jurisdiction; not always allowed)
## 3. Reading a claim
A claim is a single sentence with a defined structure:
```
[Preamble], comprising:
[Element 1];
[Element 2]; and
[Element 3].
```
The preamble may be limiting or non-limiting (case-by-case interpretation).
The body lists elements that must ALL be present for infringement.
### "Comprising" vs "consisting of"
- **Comprising:** open; additional elements OK (broader)
- **Consisting of:** closed; only these elements (narrower)
- **Consisting essentially of:** open to materials that don't affect essential features
This single word change is huge in litigation.
## 4. Claim mapping
For each candidate prior art (or competitor patent), map its claims to
your invention element-by-element.
### For prior art search
Goal: does the prior art disclose every element of my claim?
- If yes → my claim is anticipated (not novel)
- If most → may be obviousness combination with another reference
### For FTO
Goal: does my product practice every element of their claim?
- If yes → potential infringement
- If no (one element missing) → not infringing that claim
### Mapping matrix
| Your claim element | Prior art reference | Where (column/line) | Verdict |
|--------------------|---------------------|---------------------|---------|
| A | Smith 2018 | col 3:14-22 | Discloses |
| B | Smith 2018 | col 5:1-10 | Discloses |
| C | Jones 2019 | col 2:5-15 | Discloses |
| D | (none found) | | Novel |
If you find D for everything, the invention is novel. If you find a single reference disclosing all elements, the claim is anticipated.
## 5. Doctrine of equivalents
A claim covers not only literal infringement but also "equivalents":
- Substantially the same function
- In substantially the same way
- To achieve substantially the same result
This expands claim scope and complicates FTO analysis. Don't assume "we use Y instead of X" avoids infringement.
## 6. Landscape analysis dimensions
When mapping a landscape:
### By owner
- Who has filed most? (likely category leaders)
- Who's recently entered? (new threats)
- Who's exited? (signal about market)
- Who's been acquired (and where did patents go)?
### By technology cluster
- Group patents by CPC/IPC
- Find density (crowded areas) + sparseness (white space)
- White space = opportunity (or area no one's interested)
### By recency
- Filings increasing? (hot area)
- Filings declining? (mature or dying)
- Recent priority dates = recent invention
### By geography
- Where is filing concentrated?
- Where is product activity vs filing?
- National advantages (e.g., heavy biotech filing in EU)
### By claim type
- Method vs system vs composition mix
- Tells you about commercialization model
### By assignee transitions
- Inventor changes (talent moves)
- Assignee transfers (acquisitions, licensing deals)
## 7. White space identification
A white space candidate:
- No or few patents in a CPC subgroup relevant to your invention
- Existing patents are old (filing date > 10 years ago, expiration close)
- Existing patents have narrow claims (room to file broader)
- Existing patents have weak prior art behind them (could be invalidated)
White space is not always opportunity:
- Maybe no one cares (no market)
- Maybe it's not patentable (subject matter or obviousness)
- Maybe it's covered by older art too broad to easily find
Validate with market evidence + commercial diligence.
## 8. Patent strength signals
For each patent, assess:
- **Claim breadth** — narrow claims are weaker
- **Prosecution history** — many narrowing amendments = weaker
- **Citations** — highly cited = important
- **Family size** — large international family = strategic priority
- **Continuation activity** — ongoing prosecutions = active strategy
- **Litigation history** — survived challenges = strong; lost at PTAB = weak
- **Assignee** — sophisticated patentee (big tech, NPE) = more enforceable
A strong patent in an area you operate in is a real risk.
## 9. Visualizing the landscape
Useful visualizations:
### Filing timeline
- Year on x-axis, count on y-axis
- Cluster by assignee
- Shows market entry, hot periods
### Bubble chart
- X: filing year; Y: # citations; Size: family size
- Identifies dominant patents
### Network graph
- Patents as nodes; citations as edges
- Shows knowledge flow + key documents
### Heatmap
- CPC subgroups × assignees
- Color = patent count
- Identifies hotspots
## 10. Landscape outputs for strategy
A useful landscape report includes:
1. **Executive summary** — top 3 takeaways
2. **Market context** — why this area, why now
3. **Key players** — top 5-10 by patent activity
4. **Technology subareas** — clusters with hot/cold designation
5. **White space candidates** — areas thin on art
6. **Top risk patents** — for FTO purposes
7. **Recommended next steps** — file here, design-around here, license here
## 11. Common pitfalls
- **Counting patents = importance.** Some big assignees file noise; some small assignees file critical IP.
- **Ignoring patent quality.** Strong vs weak patents have different risk profiles.
- **Focus only on issued patents.** Pending applications are tomorrow's grants.
- **Forgetting continuations.** Active continuations expand scope over time.
- **Landscape without market context.** Patents without business case = expensive paperwork.
- **One-time landscape.** Refresh quarterly; landscapes shift.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |