Context cards

A React component in Beamline's AI & agents category.

The knowledge an agent retrieved before it answered: one card per chunk with what it is about, the text it read with the search words marked, how relevant it was and where it came from, filterable by source; optionally each chunk can be left out of the next answer.

Use it for

  • Showing what a retrieval step found, so people can judge an answer's grounding.
  • Letting people leave a retrieved chunk out and ask again.

Not for

  • Numbered citations inside the answer text (citations).
  • Files the person attached (attachment).
  • A search results page (a list of item rows).

Anatomy

  • header: label, "8 chunks from 5 sources · n used in the answer"
  • source filter (chip group, All + one per source with counts)
  • card: title, relevance meter (segments + word, hatched when none), text with search words marked (a tint that never splits a word), Show more
  • source line: type glyph (or connector icon), name (link), location
  • Use in answer check (when selectable)

Variants

  • layout: grid (two columns from 640px), list (side panels)

States

  • loading (skeleton cards, aria-busy)
  • empty (emptyLabel)
  • filtered by source
  • text clamped / expanded
  • relevance not scored
  • left out of the answer (unchecked; title and text step down one ink, never faded below 4.5:1)

Keyboard

  • Arrow keys choose the source filter (radio group)
  • Show more / Show less is a button with aria-expanded
  • Source names are links; Use in answer is a checkbox

Motion

None; the filter chips animate their own check (ChipGroup).

Props

PropTypeDefaultDescription
chunks (required)ContextChunk[]—{ id, title, text, source: { name, url?, location?, kind?, icon? }, score?: number | null }
querystring—Words of 3+ letters are marked in every chunk.
selected / defaultSelected / onSelectedChangestring[]—Chunks used in the answer; any of these turns the checks on.
onOpenSource(chunk) => void—
layout / previewLines"grid" | "list" / number—
loading / emptyLabel / labelboolean / ReactNode / string—
style / refCSSProperties / native Ref—
classNamesContextCardsClassNames—

Built from

Dependencies

lucide-react, radix-ui

Import

import { ContextCards } from "@/components/context-cards/context-cards";

Get it

Context cards is part of Beamline: 190 React components and screens your coding agent (Claude Code, Codex, Cursor or any MCP client) installs into your app through Beamline's MCP server, as a ready-built package or as plain React source you can change. $49 one payment (regular $200), no subscription, a year of updates, one licence for your whole team. Get Beamline · Connect your agent

More in AI & agents

  • Approval card: The agent stops and asks before it acts: what it wants to do, exactly what that changes, any choices it needs, and Approve or Reject; both run your…
  • Attachment: A file tile: preview or typed glyph with the type named, name, size, upload progress, error with retry, and a keyboard-reachable remove.
  • Chat thread: People talking to people: messages grouped by person and day, reactions, delivery and read receipts that glide to the newest message read, who is typing…
  • Citations: Inline numbered citations that preview their source on hover or focus, plus a numbered source list; streamed answers flow in at an even pace with the…
  • Conversation: The scrolling column of a conversation: follows new content while you are at the bottom, stops when you scroll up to read, and offers a way back; shared…
  • Flowchart: A workflow as steps on a dotted canvas: a trigger, conditions that branch into paths side by side, actions and waits, read top to bottom; during a run…
  • Image generation: The frame an AI image is generated into: a field of brand-hued pixels breathes with the progress while the model works, then the image arrives —…
  • Message: One turn in a conversation with an agent: the person's words in a quiet bubble, the agent's answer on the page with its orb mark, trace, tool calls and…
  • Plan: An agent's plan: numbered steps with live status, progress, and Approve / Edit / Reject when it needs sign-off; approving shows progress, and a refusal…
  • Prompt bar: The agent composer: a field that grows, @ to add a source, / for a command, a model picker, files and dictation, and one button that sends with progress…
  • Reasoning: The agent's trace, fed by your stream: a shimmering label and the seconds so far while it works, then "Thought for N seconds"; steps that check off…
  • Recommendation card: One suggestion from the agent that the person may take or leave: what it suggests, why, the expected effect and how sure it is (a meter and a word…
  • Response: An assistant's answer as streamed markdown in the system's reading type: bursty tokens flow in at an even pace, each new word fades in through a short…
  • Selection actions: Select a passage and hand it to the agent: a bar over the selection with quick actions, your own instruction and More; while the agent works it says so…
  • Task rows: Background work an agent is doing, one row per task: status as a glyph and a word, what it produced, how long it has run or took, a progress bar while…
  • Tool call: One tool call in an agent's run: status, what it did and a short result on one quiet row that opens to the exact input and output; stops for permission…