liminal-ai
Official@liminal-ai
Offers structured engineering methodologies for software specification, architectural design, code review, and full-stack development lifecycle management.
Agent Skills by liminal-ai
Showing 40 vetted skills indexed across 3 GitHub repositories.
liminal-aiCodex-agent-sdk
Plan, design, implement, debug, and extend autonomous agents with the Codex Agent SDK.
liminal-aishadcn
Manage shadcn/ui components and presets across projects.
liminal-aidoc-generation
Generate wiki-style documentation from AST-aware structural analysis of codebases.
liminal-ainext-best-practices
Guide Next.js App Router projects with centralized best practices for file conventions and data patterns.
liminal-aifrontend-design
Generate production-ready frontend components from design briefs in HTML/CSS/JS, React, or Vue.
liminal-aivercel-react-best-practices
Apply React and Next.js performance best practices to optimize codebases.
liminal-aitailwind-design-system
Centralize design tokens, components, and patterns in Tailwind CSS v4.
liminal-aiskill-development
Create standardized Claude Code plugin skills with SKILL.md frontmatter and organized resources.
liminal-aiplugin-settings
Centralize per-project plugin configuration using YAML frontmatter in local files.
liminal-aiplugin-structure
Guide Claude Code plugin design with directory, manifest, and auto-discovery conventions.
liminal-aihook-development
Implement Claude Code plugin hooks to enforce policies and validate tool usage.
liminal-aimcp-integration
Connect Claude Code plugins to MCP servers via .mcp.json or plugin.json.
liminal-aiagent-development
Design Claude Code plugin agent configurations with frontmatter and system prompts.
liminal-aireview-synthesis
Consolidate multi-pass code review findings into a verified single report.
liminal-aiperformance-review
Analyze code changes to identify performance bottlenecks and deliver prioritized optimization plans.
liminal-aifull-code-review
Orchestrate six parallel specialized code reviewers into a single verified report.
liminal-aistandard-review
Review code changes against project conventions and detect logic issues.
liminal-aisilent-failure-review
Audit error handling paths to identify and document silent failures.
liminal-aitest-coverage-review
Identify behavioral gaps in test suites across code paths.
liminal-aiarchitecture-review
Analyze code changes against project architecture and boundaries.
liminal-aisecurity-review
Identify exploitable security vulnerabilities in code and report prioritized findings in Markdown and JSON.
liminal-aiclaude-agent-sdk
Build and orchestrate Claude-powered agents with a typed multi-language API.
liminal-ailss-story
Generate a story-sized functional specification with flows and acceptance criteria.
liminal-ailss-tech
Embed architectural decisions and test mappings into functional story text.
Frequently Asked Questions About liminal-ai
FAQPage SchemaWhat specific engineering tasks does the Liminal Spec methodology enable?▼
It enables the systematic decomposition of product requirements into technical design documents, architectural boundary definitions, and executable development stories. This methodology ensures traceability from initial epic creation through to implementation, test mapping, and final verification of software features.
Which technical personas benefit most from these development capabilities?▼
Software architects, lead engineers, and full-stack developers benefit from these capabilities. The system is designed for teams requiring rigorous documentation standards, consistent architectural enforcement, and high-fidelity code reviews across complex, multi-story development projects.
What are the primary prerequisites for integrating these development plugins?▼
Integration requires a project environment configured for Claude Code, with specific support for MCP servers and YAML-based frontmatter configuration. Projects should utilize standard directory structures and manifest files to enable the auto-discovery and hook enforcement mechanisms provided by the plugin ecosystem.