Bootstrap, guide, validate, run, resume, and inspect reusable engineering automation across repositories with ordered Claude Code, Codex, Cursor, or other headless CLI runners. Use when a project needs a portable multi-agent automation contract, when a designer describes a repeatable design-to-engineering automation in plain scenario language (for example building ready-for-dev Figma components into Storybook) and needs a guided plain-language setup, when an existing `.agent-automation/config.json` task should run or resume, when runner fallback and completion proof must be auditable, or when adapting one automation workflow to a repository with different build, test, artifact, framework, or domain-specific skill requirements. Keep project-specific behavior in the project contract and companion skills instead of hardcoding it into the shared runner.
195mo agoReviewMountCreate a free account to mount
This project uses Spectra for Spec-Driven Development(SDD). Specs live in openspec/specs/, change proposals in openspec/changes/. discuss? → propose → apply ⇄ ingest → archive Changes can be parked（暫存）— temporarily moved…
195mo agoReviewMountCreate a free account to mount
Extract a reusable design-system specification from UI screenshots/images, Figma URLs or exports, Figma Variables, existing app/project folders, or prototype code. Use when Codex must produce evidence-backed design principles, design elements, token architecture, component inventory, component token specs, anti-AI style constraints, static HTML documentation for developers, cross-agent handoff guidance for Claude Code/Cursor/Codex, and a checkpoint before any product implementation.
195mo agoReviewMountCreate a free account to mount
Build or update token-backed Storybook foundations, shared UI components, and stories from an extracted design-system package. Use after design-system-extractor, or when Codex must read design-system Markdown and token files, automatically trace original design sources such as Figma URLs/nodes, UI images, rendered routes, and frontend folders, infer component dependency order so core primitives are built before composed components, map component specs into a product repo, create or update Storybook docs, plan component batches, and verify implementation with the bundled Figma export addon without bypassing tokens.
195mo agoReviewMountCreate a free account to mount
Create PRD-led product prototypes in Storybook with docs, deterministic data, typed UI Flow route and transition metadata, and prototype stories. Use when turning a product idea into PRD, UI Flow, Data Spec, Acceptance Criteria, and a clickable Storybook prototype; when scaffolding a new prototype folder; or when validating prototype docs and metadata across React Storybook projects.
195mo agoReviewMountCreate a free account to mount
Replace mock adapters with real API clients, auth/session, cache, storage, persistence, and environment configuration behind assembled DataSource seams. Use when a mock-mode feature needs real wiring, a handoff names a data-integration owner, or confirmed transport contracts and DTO-to-UI mapping need contract tests. Continue after frontend-product-implementation or native-app-implementation. Use Data Authority evidence, never treat fake fixtures as backend schemas, and keep unknown contracts open for the affected seam. Own data wiring only; do not change UI behavior, routes, components, or tokens.
195mo agoReviewMountCreate a free account to mount
Render a read-only, designer-readable HTML board showing what a project's design-to-engineering automation pipeline does, which stages have produced or verified their outputs, which handoffs are satisfied, blocked, or stale, and how many decisions are waiting for a designer. Use when a designer or a non-engineer needs to understand an automation pipeline without reading Markdown documents, run summary JSON, or asking an engineer; when a project needs an honest at-a-glance status of its extraction, design-system, and Storybook stages; or when confirming whether a downstream output was built before its upstream input changed. The board derives every claim from files already on disk and from the durable run summaries agent-automation writes. It never executes automation, never starts a runner, and requires no server, port, access code, or design tool.
195mo agoReviewMountCreate a free account to mount
Aggregate the automation-pipeline-board status of many projects into one read-only, designer-readable HTML overview dashboard: one card per tracked project showing its current stage, why it is stuck (a possibly-stalled run, the first unmet or stale handoff, or pending designer decisions), and a link to that project's own pipeline board. Use when a designer watches several projects that run design-to-engineering automation and needs one page answering which project is where and who is blocked, without opening each project or reading any run JSON. The dashboard is built by re-running the existing automation-pipeline-board build and render commands per project; a failing project becomes an error card instead of aborting the batch. It is a static snapshot: no server, no port, no polling, no run button, and it never modifies any tracked project.
195mo agoReviewMountCreate a free account to mount
This project uses Spectra for Spec-Driven Development(SDD). Specs live in openspec/specs/, change proposals in openspec/changes/. discuss? → propose → apply ⇄ ingest → archive Changes can be parked（暫存）— temporarily moved…
195mo agoReviewMountCreate a free account to mount
Plain text files in a repository that tell a coding agent how the project works: commands to run, conventions to follow and things to avoid. CLAUDE.md, AGENTS.md, cursor rules and skills are the common kinds.
CLAUDE.md or AGENTS.md?
CLAUDE.md is read by Claude Code. AGENTS.md is an open format that Codex, Cursor and other agents read. Many projects keep one and point the other at it.
What is a skill?
A folder with a SKILL.md that describes one capability, such as filling PDFs or reviewing code. The agent loads it only when the task calls for it.
Can I search my own team's files too?
Your agents already can, over MCP, limited to the files you're allowed to read. Searching them from this page is coming.
harrychuangSkill