Documentation Index
Fetch the complete documentation index at: https://docs.livepeer.org/llms.txt
Use this file to discover all available pages before exploring further.
File Placement Reference
Read this before writing any component or script file. Canonical placement rules derived from the component framework and script framework governance specifications.
Components
Root: snippets/components/
Categories (ask these questions in order)
| Question | If yes | Folder |
|---|
| Is it a single visual atom with no children, no fetching, no arrangement? | element | elements/ |
| Does it hold, group, or spatially arrange other components? | wrapper | wrappers/ |
| Does it take content and present it in a formatted way? | display | displays/ |
| Is it part of the page’s outer structure, typically used once? | scaffolding | scaffolding/ |
| Does it fetch, embed, or bind to external/third-party data? | integrator | integrators/ |
| Is it a non-component config/theme object? | config | config/ |
Sub-niches by category
- elements: a11y, buttons, callouts, icons, images, links, math, social, spacing, text
- wrappers: accordions, cards, containers, grids, lists, steps, tables
- displays: code, diagrams, quotes, response-fields, video
- scaffolding: frame-mode, heroes, portals
- integrators: blog, embeds, feeds, video-data
Placement path
snippets/components/<category>/<subniche>/ComponentName.jsx
Common mistakes
| Wrong | Right | Why |
|---|
BadgeWrapper in elements/ | BadgeWrapper in wrappers/ | Category is determined by what the component DOES, not what it is named after |
| New component in closest-sounding folder | Read the 6 questions above | Pattern-matching on name produces wrong placement. Use the decision questions |
| Component in repo root | Never | All components go in snippets/components/<category>/<subniche>/ |
@component, @type, @subniche, @status, @description, @accepts, @dataSource (if integrator)
Scripts
Root: operations/scripts/
Types (Layer 1)
| Type | Folder | What it does |
|---|
| audit | audits/ | Read-only scan, measure, report |
| generator | generators/ | Produce new files from source data |
| validator | validators/ | Enforce rules with pass/fail gate |
| remediator | remediators/ | Bulk fix or repair existing files |
| dispatch | dispatch/ | Dispatch work to other scripts or agents |
| automation | automations/ | Automated pipelines: translation, data fetching, transforms |
Concerns (Layer 2)
| Concern | Folder | What it covers |
|---|
| content | content/ | Docs pages, copy, SEO, veracity, quality |
| components | components/ | Component library, registry, CSS, naming |
| governance | governance/ | Scripts about scripts, repo structure, agent docs |
| ai | ai/ | LLM files, agent packaging, skills sync |
Placement path
operations/scripts/<type>/<concern>/<niche>/script-name.js
@script, @type, @concern, @niche, @purpose, @description, @mode, @pipeline, @scope, @usage, @policy
Before writing ANY file
- Determine the category using the decision questions (components) or type+concern (scripts)
- Check the target folder exists
- If unsure, state: “This component does [X]. That makes it a [category] because [reasoning]. Placing at [path]. Confirm?”
- Never place files based on name pattern-matching. Always reason from what the component/script DOES
Canonical sources
