AI + automation

AI Connect

FavStash is built AI-first: a hosted MCP endpoint that turns your stash, scheduler, and analytics into tools your AI can use.

Get your API key Read the install prompt

Works with

GPT / Codex

Claude

Cursor

Windsurf

OpenClawcustom MCP clients + more

Agentic setup

Teach your AI agent to retrieve your saved reels, posts, transcripts, and collections from FavStash.

Install — copy this prompt and paste it into your AI agent

The agent reads the setup guide, asks for your API key, and connects to the hosted MCP endpoint.

text

Read https://www.favstash.app/INSTALL_FOR_AGENTS.md and follow it. Ask me for my FavStash API key if you need it.

Install prompt

Get your API key

Mint a personal access token from the FavStash console. The setup prompt tells your agent to ask for it.

bash

export FAVSTASH_API_KEY="favstash_pk_YOUR_TOKEN_HERE"
codex mcp add favstash \
  --url https://mcp.favstash.app/mcp \
  --bearer-token-env-var FAVSTASH_API_KEY

Direct Codex command if you want to configure it yourself.

Try it

After connecting, use prompts like these to verify retrieval, counts, filters, and transcripts.

prompt

Use FavStash to find reels I saved this week about AI product ideas. Include links and short summaries.

Semantic search plus saved-at filtering.

prompt

Use FavStash to list my collections, then search the most relevant collection for app store rejection examples.

Collection discovery before filtered retrieval.

prompt

Use FavStash to count how many Instagram reels I saved in the last 7 days, without returning full transcripts.

Count-only inventory query.

prompt

Use FavStash to get the full transcript for the best matching saved reel about UI design tools.

Detailed item retrieval only when needed.

What gets connected

Your agent gets owner-scoped tools for semantic search, recent saves, stash counts, collections, item detail, transcripts, and URL saving.

What stays private

Agents authenticate with your personal token. Tokens can be revoked from the console, and the backend never accepts an owner id from the tool payload.

Custom MCP

Drop these into any MCP-aware client. Same endpoint, same Bearer token.

MCP endpoint

text

URL                  https://mcp.favstash.app/mcp
Authorization        Bearer favstash_pk_YOUR_TOKEN_HERE
Content-Type         application/json
Protocol             MCP over HTTP (Streamable HTTP, JSON-RPC 2.0)

The minimum any MCP HTTP client needs to connect.

Claude Desktop config

json

{
  "mcpServers": {
    "favstash": {
      "type": "http",
      "url": "https://mcp.favstash.app/mcp",
      "headers": {
        "Authorization": "Bearer favstash_pk_YOUR_TOKEN_HERE"
      }
    }
  }
}

Drop into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) and restart Claude. Look for 'favstash' under the tools menu.

Raw HTTP examples

Useful for debugging or custom clients.

bash

# Quick smoke test — list MCP tools
curl -s -X POST https://mcp.favstash.app/mcp \
  -H "Authorization: Bearer favstash_pk_YOUR_TOKEN_HERE" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Smoke test — should return the tool descriptors. Health probe (no auth): GET https://mcp.favstash.app/health

bash

# Semantic search across your stash
curl -s -X POST https://mcp.favstash.app/mcp \
  -H "Authorization: Bearer favstash_pk_YOUR_TOKEN_HERE" \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 2,
    "method": "tools/call",
    "params": {
      "name": "search_items",
      "arguments": {
        "query": "react hooks",
        "sourcePlatform": "youtube",
        "dateFrom": "2026-05-02T00:00:00.000Z"
      }
    }
  }'

Semantic retrieval with explicit filters. Use ISO timestamps for saved-at ranges.

bash

# Count-only query for agents
curl -s -X POST https://mcp.favstash.app/mcp \
  -H "Authorization: Bearer favstash_pk_YOUR_TOKEN_HERE" \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 3,
    "method": "tools/call",
    "params": {
      "name": "get_stash_stats",
      "arguments": { "since": "2026-05-02T00:00:00.000Z" }
    }
  }'

Count-only retrieval for inventory questions without returning transcripts.

bash

# Schedule a post to a connected channel
# (get connectionId from the list_channels tool)
curl -s -X POST https://mcp.favstash.app/mcp \
  -H "Authorization: Bearer favstash_pk_YOUR_TOKEN_HERE" \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 4,
    "method": "tools/call",
    "params": {
      "name": "schedule_post",
      "arguments": {
        "connectionIds": ["YOUR_CONNECTION_ID"],
        "content": "Caption goes here",
        "media": [{ "url": "https://example.com/clip.mp4" }],
        "publishAt": "2026-07-10T18:00:00Z"
      }
    }
  }'

Schedule a post to Instagram, YouTube, or TikTok. Media must be a public HTTPS URL — import anything else first with upload_media_from_url.

Here's the thing that makes FavStash different: everything is built for AI. The stash, the scheduler, the analytics — all of it is exposed through a hosted MCP endpoint, so any MCP-capable assistant can operate FavStash on your behalf, in plain conversation.

Not a chatbot bolted onto an app. The app itself, as a toolbox your AI already knows how to hold.

What your AI can do#

Once connected, your assistant gets owner-scoped access to the same 19 tools the apps use:

Search your stash for inspiration — "find the reels I saved about product launches and pull their hooks."
Save links for you — paste a URL mid-conversation, it lands in your stash with full enrichment.
Plan a content calendar — it can read what you've saved, what you've posted, and what performed, then propose a week.
Bulk-schedule across channels in one conversation — a week of posts, multiple platforms, one chat.
Pull analytics reports and suggest what to double down on — per-channel and per-post numbers, interpreted instead of just listed.

Prompts that actually work

Schedule this reel idea to Instagram and TikTok every Tuesday 6pm this month.

Which of my last 5 posts performed best and why? Pull the numbers first.

Search my stash for the "hook ideas" collection, pick the 3 strongest hooks,
and draft a TikTok script for each.

Save https://www.instagram.com/reel/... to my Inspiration collection with the
note "remix this format for the launch".

Pull my YouTube analytics for the last 30 days and tell me what to post more of.

Get connected#

1. Mint a personal access token

Go to Portal → AI Connection and create an API key. It looks like favstash_pk_... — copy it somewhere safe, because it's shown once. Keys can be named, tracked (last-used time), and revoked individually.

Minting a key is free for everyone. Tool calls require a paid plan (Doom Scroller or Creator Pro). You can choose an optional 3-day free trial once per account, so you can watch your AI work before paying. Cancelling during the trial ends access immediately. Until you subscribe, your AI gets a friendly upgrade prompt instead of your data.

2. Connect your agent (pick your flavor)

All the copy-paste configs live in the interactive setup at the top of this page — pick your client (Claude, ChatGPT, Codex, Cursor, Windsurf), grab the agentic install prompt, the MCP config, or raw HTTP examples, each with a copy button.

Anything else. Cursor, Windsurf, OpenClaw, custom clients — any MCP client works. The minimum contract:

URL              https://mcp.favstash.app/mcp
Authorization    Bearer favstash_pk_YOUR_TOKEN_HERE
Protocol         MCP over HTTP (Streamable HTTP, JSON-RPC 2.0)

3. Try it

Ask your assistant something only your stash can answer:

Use FavStash to find reels I saved this week about AI product ideas.
Include links and short summaries.

If it comes back with your actual saves, you're live.

Security model#

One user, one scope. Every token is scoped to your account. The backend enforces ownership server-side on every tool call — a client can't read or write anyone else's data no matter what arguments it sends.
Tokens are revocable. Kill any key instantly from Portal → AI Connection.
Your channel logins stay home. The AI never sees your Instagram/YouTube/TikTok credentials or OAuth tokens — it calls FavStash tools, and FavStash talks to the platforms.

The full toolbox#

Every tool, grouped and documented with key params, lives in the MCP tools reference.