DNT
Community@MathematicGuy · Internet
Coder in heat
Agent Skills by DNT
Showing 70 vetted skills indexed across 1 GitHub repositories.
lumi-ask
Answer questions by traversing a wiki knowledge graph and synthesizing cited responses.
lumi-research-watchlist
Configure scheduled research topic watchlists in YAML for automated paper discovery.
lumi-learning-reflect
Creates and updates personal reflection pages in a wiki with user-written understanding and append-only evolution logs.
lumi-migrate-legacy
Backfills missing frontmatter schema fields on legacy wiki entries after version upgrades.
lumi-edit
Applies user-directed edits to wiki pages while preserving cross-reference edge integrity and lint compliance.
lumi-research-prefill
Creates Wikipedia-backed foundation pages for stable background concepts in a wiki.
lumi-check
Runs the wiki linter, classifies findings by severity, and applies confirmed auto-fixes with a verification re-run.
lumi-help
Orient users in a Lumina wiki workspace via orientation, catalog, and framework Q&A modes.
lumi-research-setup
Configure research pack runtime by verifying Python tools and populating API-key environment files.
lumi-research-topic
Cluster existing wiki concepts and sources into cross-referenced topic pages with bidirectional graph edges.
lumi-research-discover
Discover and rank candidate research sources from arXiv, Semantic Scholar, and OpenAlex into a reviewable shortlist.
lumi-verify
Verify wiki notes against cited sources using three isolated reviewer agents.
lumi-research-watch-run
Runs one discovery pass over a research watchlist of topics and RSS feeds.
lumi-init
Initializes a Lumina wiki workspace by creating directories, seeding index and log pages, and verifying lint.
lumi-ingest
Converts raw source files into structured wiki pages with graph edges and citations.
lumi-research-survey
Synthesize grounded survey narratives from existing wiki sources, concepts, and graph edges.
lumi-reset
Deletes wiki, raw, or state files with dry-run plans and explicit confirmation.
ui-pattern
Converts a one-paragraph app idea into PRD.md and GUIDELINE.md files through a structured brainstorm conversation.
tcr-apply
Generates three additive Transparency, Control, and Recovery prompts tailored to an existing UI pattern.
prd-to-screens
Convert a PRD into a prioritized UI pattern selection and GUIDELINE.md build specification.
ux-audit
Evaluate interface designs against UX principles and generate structured audit reports.
a20
Generates course journals, worklogs, ADRs, and pull requests for VinUni A20 students.
standup
Generates daily standup report suggestions from git commits and GitHub Project items.
flow
Guides projects through 11 gated stages from idea to deployed URL using a deterministic shell checker.
Frequently Asked Questions About DNT
FAQPage SchemaWhat tasks can I accomplish with DNT's skill collection?▼
You can build and maintain a Lumina personal wiki (ingest sources, query the graph, verify citations, run lint checks), run BMad product cycles (PRDs, epics, stories, sprint planning, retrospectives), conduct systematic literature reviews across PubMed and arXiv, and design AI app interfaces using 7 UI patterns with UX audits.
Who are these skills designed for?▼
Three main personas: individual researchers building a personal knowledge graph with Lumina, product teams and engineers using BMad agents (PM, architect, developer, UX designer) for spec-driven development, and VinUni students in the A20 course needing journals, worklogs, standup reports, and PR compliance checks.
How do the Lumina wiki skills work in practice?▼
Start with lumi-init to bootstrap the workspace, then lumi-ingest to convert raw files or arXiv IDs into structured wiki pages through four review checkpoints. Use lumi-ask for cited graph queries, lumi-check for lint health, lumi-verify for source fact-checking, and lumi-edit for guided page revisions.
Are these skills free and open source?▼
Licensing is only explicitly stated on two skills: prd-to-screens and literature-review both carry MIT licenses. The remaining skills in the manifest do not declare a license, so users should check each skill's repository folder before redistribution or commercial use.
What prerequisites do these skills require?▼
Lumina skills require a Lumina wiki workspace with Bash and file read/write access; the research pack additionally needs Python dependencies and API-key environment files configured via lumi-research-setup. BMad skills operate on project repositories with markdown specs, and literature-review needs access to academic databases like PubMed and arXiv.