All skills

Turns approved game design and art direction into a runnable whitebox prototype or production build on the locked target runtime.

A StyleSeed skill that converts a human-approved UI design correction into a privacy-minimized local rule candidate, with explicit review and opt-in share steps.

Builds institution-grade research reports — industry reports, white papers, annual surveys, consulting decks, and arXiv papers — covering structure, prose, charts, and page-level print QA.

Enforces a discover → inspect → plan → query → validate workflow for any data question against a ktx-connected database, with a deep set of correct-SQL heuristics.

A ktx-specific skill that mines historic SQL query shards for recurring cross-table analytical intents and emits verified pattern evidence.

Maps dbt schema.yml models, sources and tests into ktx semantic-layer overlays and column notes without inventing physical columns.

A playbook for using the GitHub CLI gh-stack extension to split multi-part work into reviewable stacked branches and PRs, safely and non-interactively.

Read-only audit of the skills, MCP servers, hooks, plugins, and credentials already installed across Claude Code, Codex, OpenClaw, and Cursor, with a structured risk briefing.

Drive the Codeman session manager over its HTTP API to spawn, prompt, wait on, and clean up worker sessions across any installed agent CLI.

A strict automation that reproduces Slack-triaged bugs twice through the real UI and opens a draft PR only after before-and-after proof.

Classifies a single Slack bug report, dedupes against your issue tracker, and posts exactly one verdict back in the source thread.

A code-first skill for installing, running, and fine-tuning RWKV — the RNN+Transformer hybrid with O(n) inference and constant-memory infinite context.

Decides which model, effort level, and cascade shape each subagent gets, routing on measured cost per completed task instead of per-token price.

Generates WeChat Official Account cover images and in-article illustrations by matching the article's title and content to reusable style presets.

Places risk-approved orders one at a time through Alpaca or Interactive Brokers, never without an explicit typed yes plus Claude's own Allow once.

Designs production-grade multi-agent workflows with the right pattern, explicit handoff contracts, failure handling, and cost/context budgets.

Scans AI coding-agent caches (.claude, .cursor, .antigravity, .openclaw) or any folder you name, then turns scattered experiment logs into PaperOrchestra-ready idea.md and experimental_log.md.

Turns your idea notes, experiment logs, LaTeX template and conference guidelines into a strict JSON paper outline with figure, literature-search and section plans.

An iterative LaTeX paper refinement loop that simulates peer review and only keeps revisions that provably raise the score.

Discovers candidate papers via web search, verifies them through Semantic Scholar with Crossref/OpenAlex cross-checks, then builds refs.bib and drafts the Introduction and Related Work.