Danny avatar

Danny

Community

@djfksjd · korea

18Followers
|
16Public Repos
|
1Published Skills

Korean government and public-sector startup grant discovery skill that crawls K-Startup, Bizinfo, NIPA, KOCCA, and SMTECH to match projects against funding eligibility.

Skills Distribution
DomainBusiness, Fi...Government Grant &.. (40%)Eligibility Assess.. (25%)Public Portal Web .. (20%)Report Generation .. (15%)

Agent Skills by Danny

Showing 1 vetted skills indexed across 1 GitHub repositories.

Frequently Asked Questions About Danny

FAQPage Schema
What tasks can I accomplish with the ir-search skill?

It performs exhaustive searches of Korean government and public startup support programs—funding, commercialization capital, office space, R&D, vouchers, and competitions—then classifies matches into three eligibility tiers and produces a report with verified deadlines and requirements from original postings.

Who is the target user for ir-search?

Korean startup founders, project teams, and entrepreneurs seeking new-item grants, R&D funding, or incubation programs. It is not for operating small-business owners seeking merchant subsidies or policy loans—those requests route to the separate sole-search skill instead.

Which data sources does ir-search crawl?

It crawls five Korean public portals: K-Startup, Bizinfo (기업마당), NIPA, KOCCA, and SMTECH. Announcements are matched against the project profile in the current working folder, with eligibility and deadlines verified against original source text.

How does ir-search handle repeat or follow-up searches?

When a user asks for re-investigation or newly posted programs on a previously researched project, it runs in diff mode—an incremental rescan that surfaces only announcements added since the last search, avoiding redundant full crawls.

How does ir-search decide between startup grants and small-business support?

Request purpose overrides user identity: new-item startup or R&D searches use ir-search, while operating store-owner subsidies use sole-search. If signals are mixed (e.g., 'online seller grants'), the skill asks the user once to clarify intent.