Xinyi_Zheng
Community@zniihgnexy · Bristol
#hate work
Agent Skills by Xinyi_Zheng
Showing 49 vetted skills indexed across 1 GitHub repositories.
zniihgnexygithub
Triages GitHub repositories, pull requests, and issues through a connected GitHub app.
zniihgnexygh-fix-ci
Inspect failing GitHub Actions PR checks and extract failure log snippets with gh.
zniihgnexygh-address-comments
Inspect unresolved GitHub pull request review threads and implement selected fixes locally.
zniihgnexyfigma-generate-design
Builds or updates Figma screens from code using design system components and tokens.
zniihgnexyfigma-create-new-file
Creates a new blank Figma design or FigJam file via the create_new_file MCP tool.
zniihgnexyfigma-generate-library
Builds Figma design systems with variables, components, and theming from a codebase.
zniihgnexyfigma-use
Executes JavaScript in Figma files via the Plugin API to create and edit design nodes.
zniihgnexyfigma-code-connect
Creates parserless Figma Code Connect template files mapping Figma components to code snippets.
zniihgnexyfigma-create-design-system-rules
Generates project-specific design system rules for Figma-to-code workflows.
zniihgnexyanalysis-campaign
Orchestrates coordinated follow-up experiment campaigns for ablations, robustness checks, and failure analysis.
zniihgnexybaseline
Establishes verified reference baselines through attach, import, reproduce, or repair routes.
zniihgnexyintake-audit
Audits and trust-ranks existing research assets to select the next workflow anchor.
zniihgnexyreview
Audits paper drafts skeptically and produces review reports, revision logs, and experiment TODO lists.
zniihgnexyoptimize
Manages candidate briefs, optimization frontiers, and branch promotion for algorithm-first research quests.
zniihgnexyexperiment
Executes and records auditable main experiment runs against an accepted baseline.
zniihgnexydecision
Records evidence-backed go, stop, branch, and finalize decisions for research quests.
zniihgnexyfigure-polish
Generates and revises academic-style Matplotlib figures through a render-inspect-revise workflow.
zniihgnexyfinalize
Consolidates research quest evidence into final claims, limitations, recommendations, and handoff packets.
zniihgnexyrebuttal
Convert reviewer feedback into experiments, manuscript revisions, and a structured rebuttal response package.
zniihgnexywrite
Drafts evidence-grounded research papers and reports with LaTeX venue templates and claim-evidence tracking.
zniihgnexyscout
Frames research quests by scouting literature, evaluation contracts, and baseline candidates.
zniihgnexyskill-creator
Create and validate Codex skill directories with SKILL.md templates and helper scripts.
zniihgnexyplugin-creator
Scaffold Codex plugin directories with plugin.json manifests and marketplace.json entries.
zniihgnexyskill-installer
Install Codex skills from curated GitHub lists or arbitrary repo paths.