Taku Matsumura
Community@matsumuratk
Taku Matsumura's skill registry covers web research pipelines, GTM/GA4 tag testing, image and diagram generation, PPTX editing, and Streamlit dashboard design.
Agent Skills by Taku Matsumura
Showing 15 vetted skills indexed across 2 GitHub repositories.
matsumuratkresearch
Generates research outlines and field definitions for a topic using web search.
matsumuratkgenshijin
Compresses Japanese AI responses into terse keyword-style output to reduce token usage.
matsumuratkresearch-deep
Launches parallel research agents from an outline file to produce validated structured JSON outputs.
matsumuratknano-banana
Generate and edit images via the Gemini CLI nanobanana extension.
matsumuratkresearch-brief
Generates structured web research reports from a research_brief.md specification file.
matsumuratkresearch-add-fields
Appends user-confirmed field definitions to an existing research fields.yaml file.
matsumuratkresearch-report
Generates a markdown summary report from deep research JSON results via a Python script.
matsumuratkdrawio
Generates draw.io diagrams as native .drawio XML files with optional PNG, SVG, or PDF export.
matsumuratkpptx
Create, edit, and validate PowerPoint .pptx presentations via XML manipulation and pptxgenjs.
matsumuratkbrowser-use
Automates browser navigation, interaction, and data extraction via a persistent CLI daemon.
matsumuratkstreamlit-design-digitalagency
Builds and reviews Streamlit and HTML dashboards following the Digital Agency dashboard design guidebook.
matsumuratkadd-changeset
Create changeset markdown files documenting user-facing changes for release notes.
matsumuratkadd-competitor
Researches an AI visibility tool and inserts a competitor entry into the directory codebase.
matsumuratkgtm-make-case
Generate GTM test case files from simplified instruction files.
matsumuratkgtm-test
Execute predefined GTM preview mode test cases and validate dataLayer and GA4 events.
Frequently Asked Questions About Taku Matsumura
FAQPage SchemaWhat tasks can I accomplish with matsumuratk's skills?▼
You can run structured web research from outline to markdown report, generate and edit images via Gemini, create drawio diagrams and PPTX decks, build Streamlit dashboards per Japan's Digital Agency design guide, automate browser interactions, and validate GTM/GA4 tagging with predefined test cases.
Who is the target user for these skills?▼
Marketers, analysts, and developers managing GA4/GTM implementations, content producers needing images, thumbnails, diagrams, and slide decks, researchers compiling benchmark or technology-selection reports, and engineers maintaining changesets or an AI tool competitor directory.
How does the research workflow operate in practice?▼
Start with the research skill to generate an outline, extend it with research-add-items and research-add-fields, then run research-deep to launch per-item deep investigation, and finish with research-report to summarize all fields into a markdown report, skipping uncertain values.
Are these skills open source and what do they cost?▼
Licensing varies per skill. The pptx skill is explicitly proprietary with terms in its LICENSE.txt file. Other skills carry no stated license in the manifest, so review each skill's repository before commercial redistribution or modification.
What prerequisites or dependencies do these skills require?▼
Image generation requires the Gemini CLI via Bash, browser-use requires its browser automation runtime, and research skills rely on WebSearch, WebFetch, and Task capabilities. GTM testing requires predefined test case files and a GTM preview-mode environment with GA4 instrumentation.