recipe-patent-landscape
Skill1 starsChanged 4 months agoMITView on GitHub
Patent-landscape analysis for a technology area — scoped dual-database search, key-player identification, full-corpus filing analytics, and CPC-versus-year white-space detection. Trigger when the user asks to map a technology space, identify who patents in an area, or report filing trends and white space.
In this repository34 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- flowleap-academicSkill
- flowleap-authSkill
- flowleap-citationSkill
- flowleap-keysSkill
- flowleap-legalSkill
- flowleap-nplSkill
- flowleap-opsSkill
- flowleap-patentSkill
- flowleap-patstat-graphSkill
- flowleap-patstatSkill
- flowleap-sharedSkill
- flowleapSkill
- flowleap-toolsSkill
- flowleap-usptoSkill
- persona-ip-analystSkill
- persona-patent-attorneySkill
- persona-researcherSkill
- persona-startup-founderSkill
- recipe-academic-literature-reviewSkill
- recipe-audit-reportSkill
- recipe-claim-analysisSkill
- recipe-claim-draftingSkill
- recipe-custom-dashboardSkill
- recipe-find-betterSkill
- recipe-freedom-to-operateSkill
- recipe-infringement-chartingSkill
- recipe-invalidity-analysisSkill
- recipe-invention-disclosureSkill
- recipe-maintenance-feesSkill
- recipe-office-action-responseSkill
- recipe-patent-to-reportSkill
- recipe-prior-art-searchSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 0 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 21
Write the scoping queries yourself (method in flowleap-patent): extract the area's specific noun phrases, keep at least one discriminating …
- Very specific: line 33
Verify any CPC code against the official scheme before scoping on it — flowleap patstat query on flowleap.cpc_scheme (see flowleap-patent).
- Very specific: line 47
If one office answers provider_keys_required, its patent-data key is missing: map the live office in full, label the landscape as covering …
- Very specific: line 96
When tallying players or counts from the search results, collapse to one entry per patent family so multi-jurisdiction filings are not …
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Recipe: Patent Landscape Analysis
Map the patent landscape for a technology area, identifying key players, trends,
and gaps. Each database uses its own query syntax — see flowleap-uspto for the
USPTO Lucene grammar.
In a chat client with the FlowLeap connector or in the FlowLeap Patent AI app, call the tools named in the flowleap-shared tool table instead of these commands, and never run a flowleap command there. The steps stay the same.
Steps
Step 1: Define Search Scope
Write the scoping queries yourself (method in flowleap-patent): extract the
area's specific noun phrases, keep at least one discriminating term (a CPC
class alone maps a continent, not a landscape), and probe the count — a
landscape query legitimately lands broader than a novelty query, but a count
in the tens of thousands still means the scope is the technology area, not
the landscape:
# Count probe (read `total` from the payload), then adjust scope
flowleap --json tools run search_patents query='<self-written CQL>' range=1-1 details=falseVerify any CPC code against the official scheme before scoping on it —
flowleap patstat query on flowleap.cpc_scheme (see flowleap-patent);
group titles carry the specific technology, the 4-char class only the
headline.
Done when you have a probed EPO CQL query and a USPTO ODP query for the area.
Step 2: Broad Patent Search
flowleap --json patent search --query "<CQL from step 1>" --limit 50
flowleap --json uspto search --query "<ODP Lucene from step 1>" --limit 50If one office answers provider_keys_required, its patent-data key is missing:
map the live office in full, label the landscape as covering that office only
because of a missing key — never as the shape of the field — and ask for the
free key at the end. flowleap patstat aggregates stay available keyless, but
they are twice-yearly snapshot counts, not a substitute for live search. See
flowleap-keys.
Done when both databases have returned their result sets, or the gated office is named as an open missing-key gap.
Step 3: Corpus Analytics
# Filing trends by year, country and CPC breakdowns, top assignees
flowleap --json analytics --keyword "<technology>" --date-from 2015-01-01
flowleap --json analytics --cpc <cpc-prefix> --country US --date-from 2020-01-01Step 4: Identify Key Players
Scope the Step 1 query per applicant with pa= — wildcards catch name
variants (pa=GOOGLE* catches "Google LLC" and "Google Inc"), and
subsidiaries file separately (Google also files as Alphabet, DeepMind,
Waymo):
flowleap --json patent search --query 'pa=<ASSIGNEE>* AND <discriminating terms from step 1>' --limit 30Step 5: Check Recent Activity
patent search returns relevance-ranked hits; ops search --cql adds CQL
date-range filtering (pd>=2024) for a time-sliced view the ranked search does
not expose:
flowleap ops search --cql "ti=<technology> AND pd>=2024" --start 1 --end 50Step 6: Flag White Space
Cross the analytics CPC breakdown against the filing-year trend to flag subclasses that are sparse or declining while neighbours grow. Done when at least one sparse/declining CPC subclass (or a confirmed absence) is identified.
Output
A dataset segmented by database, applicant, and filing date, plus corpus-level
trend charts (filings per year, top assignees, CPC and country distributions).
When tallying players or counts from the search results, collapse to one entry
per patent family so multi-jurisdiction filings are not double-counted; the
corpus analytics figures are aggregate backend counts, reported as returned.
Visual deliverable
To turn Step 3's filing-trend numbers and Step 6's white-space finding into a
shareable HTML dashboard, follow this recipe's analysis through to the end,
then render it with recipe-custom-dashboard — its landscape white-space
template (CPC × year heatmap) is built for this recipe's final step, and its
filing-trends template covers Step 3's year-over-year counts. Analysis
logic stays here; the dashboard skill only owns presentation.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

