The hosted data surface
for AI agents.

Stop making agents rebuild fragile tables in chat. Give your agents one endpoint to publish research, logs, and scraped data into a live, persistent, and embeddable data grid.
No database. No frontend. No schema setup.

$ npm i -g instadash
✓ MCP-native✓ Token-efficient✓ Human-in-the-loop
~/projects/au — bash
$
instadash · you/q2-reportlive
regioncategoryrevenuegrowthtrend
APACSaaS$2.4M+18%
AMEREnterprise$5.1M+31%
EMEASMB$1.8M+12%
LATAMStartup$0.6M+44%
01the problem

Agents shouldn't render grids.
Your context window isn't a spreadsheet.

AI agents generate structured data constantly, but displaying it inside a chat window is slow, wasteful, and ephemeral. Markdown tables get truncated, formatting breaks, and the data disappears when the session ends. Instadash gives that data a permanent home that both humans and agents can use.

BEFORE — agent renders in chat
Claude — conversation
Here's the comparison you asked for:
| Company  | Revenue | Growth | Cap    | Mo...
|----------|---------|--------|--------|------
| Acme     | $2.4M   | +18%   | $120M  | [tr…
| Globex   | $5.1M   | +31%   | $340M  | [tr…
| Initech  | $1.8M   | +12%   | $89M   | [tr…
| Umbrella | $0.6M   | +44%   | $210M  | [tr…
⚠ Table truncated — 4 columns hiddenSession will clear on refresh.
AFTER — agent publishes to Instadash
~/projects — bash
$ instadash push --name comparison
✓ 10 rows · 8 cols inferred
✓ persistent · queryable · private
✓ instadash.io/you/comparison
02pipeline

From agent output to live grid in seconds.

Your agent produces JSON. One push creates a persistent grid, queryable API, and LLM context — accessible to humans and agents alike.

01→
PIPE

Send JSON, CSV, NDJSON, logs, or agent output from anywhere.

output
cat data.json | instadash push
02→
INFER

Columns, types, dates, links, images, and numbers are detected automatically.

output
schema inferred
03→
VIEW

A live grid URL — humans can search and filter, agents can query via MCP or API.

output
instadash.io/you/sales-q2
04
USE

Every push ships llms.txt, llms.md, a queryable API, and a versioned snapshot. One endpoint, reusable forever.

output
llms.txt · llms.md · /api · Time Machine
03integrations

Give your agents a real data surface.

Built for MCP workflows. Agents publish structured data and return a live grid URL — queryable, persistent, embeddable.

one-liner — mcp
// claude_desktop_config.json
"instadash": {
  "command": "npx",
  "args": ["-y", "instadash-mcp"],
  "env": { "INSTADASH_KEY": "sk_…" }
}
HTTP 200 · response.json87 ms
{
  "status": "processing",
  "handle": "au-housing",
  "slug": "sales-q2",
  "version": 1,
  "resources": {
    "grid": "instadash.io/au-housing/sales-q2",
    "embed": "instadash.io/embed/sales-q2",
    "api": "instadash.io/au-housing/sales-q2/rows",
    "llms_txt": "instadash.io/au-housing/sales-q2/llms.txt",
    "llms_md": "instadash.io/au-housing/sales-q2/llms.md",
    "mcp": "instadash://grids/au-housing/sales-q2"
  }
}
MCP TOOLS
instadash_pushinstadash_readinstadash_listinstadash_get_schemainstadash_list_versions
RECOMMENDED AGENT BEHAVIOUR

When an agent needs to show structured data, search results, scraped output, logs, comparisons, extracted entities, or research tables — publish the data to Instadash and return the live grid URL instead of rendering a large table in chat.

04embeddable

A live grid inside any workflow.

Every grid includes an embed URL. Drop it into internal tools, docs, portals, or dashboards.

THEME
instadash · test-builder/comparison · live embedlive
EMBED CODE
<iframe
  src="https://instadash.io/test-builder/comparison?embed=true&theme=dark"
  width="100%" height="420" frameborder="0"
></iframe>
05use cases

Built for structured agent output.

Any agent that produces structured data has a home in Instadash.

Claude + MCP
+instadashFEATURED RECIPE2 / 6
Claude → Instadash in 30 seconds.

Paste one URL into Claude.ai, Cursor, or Windsurf. Your grids appear in the tool list. Seven tools, OAuth, zero config.

Paste MCP URL→
OAuth sign-in→
Tools appear→
Agent runs
claude desktop · mcp
# paste this URL into your client
> https://mcp.instadash.io
 
✓ instadash_list      list workspace grids
✓ instadash_push      create / update a grid
✓ instadash_read      paginate rows
✓ instadash_edits     read HITL human edits
✓ instamesh_search    search the public mesh
RESEARCH
Research agents
Competitor research, source lists, pricing tables, market maps, extracted claims.
CODING
Coding agents
Logs, test results, API responses, migration output, debugging traces.
OPS
Ops agents
Incidents, queues, sync runs, error reports, webhooks, system snapshots.
SALES
Sales agents
Prospect lists, account research, enrichment output, lead scoring, outreach queues.
SUPPORT
Support agents
Tickets, themes, product feedback, account issues, customer records.
06two surfaces

One push. Two surfaces.

Your agent finds it. You review it. Your agent acts on it.

GRID · FOR HUMANS

See the data.

The review surface for agent output. Search, filter, and inspect rows — then hand back to your agent with confidence.

  • ›Search and filter agent output
  • ›Deep links to share exact views
  • ›Track every agent push with Time Machine
  • ›Schema inferred automatically
cat data.json | instadash push
AGENT · FOR LLMS

Reuse the data.

Every grid exposes API endpoints and LLM-ready context. Agents push data in, humans review it, agents query it back — all from the same URL.

  • ›llms.txt and llms.md on every grid
  • ›Queryable API endpoints
  • ›MCP-native: push and read via tool calls
  • ›Token-efficient — data lives outside the chat context
agent.tool('instadash_push', { data: rows })
07private by default

Agent output should not leak.

Grids are private by default. Use Instadash for internal logs, research output, scraped data, and operational snapshots — without making every grid public by accident.

  • ✓Private grids by default — nothing is public until you say so
  • ✓Public sharing when enabled — one toggle, one URL
  • ✓Read-only embeds — viewers can inspect, not edit
  • ✓API-key protected access — every request requires your key
  • ✓TTL support — set grids to auto-expire after a fixed period
  • ✓Delete any grid instantly — no retention after you remove it
08pricing

Start free. Pay as you scale.

Free forever. Private by default. Upgrade for higher limits, longer retention, and agent workflows. No per-seat pricing.

free
$0/forever

Try Instadash. Private by default, no credit card.

  • ›3 grids
  • ›1k rows / grid
  • ›7-day version history
  • ›CLI upload
  • ›Private by default
most popular
builder
$19/mo

For builders shipping real workflows.

  • ›Unlimited grids
  • ›Private grids
  • ›100k rows / grid
  • ›30-day version history
  • ›CSV, JSON, NDJSON, logs
  • ›API endpoints
  • ›LLM context files
  • ›Embeddable grids
agent
$79/mo

For agents, automations, and scheduled data flows.

  • ›Unlimited grids
  • ›5M rows / grid
  • ›Hosted MCP server
  • ›Grid search API
  • ›Webhooks
  • ›Scheduled refresh
  • ›Team sharing
  • ›Higher API limits
platform
Custom

For teams that need control over infra.

  • ›BYO Cloudflare account
  • ›Custom domain
  • ›SSO + audit logs
  • ›Private mesh sub-network
  • ›Self-hosted Worker
  • ›SLA
────────────────────────────────────────ready────────────────────────────────────────

Your agent has a place to put its data.

Free to start. No database. No frontend. Just an endpoint for your agent.