# STEALCONTENT

> A library of social posts that already worked, cut by vertical, so an agent can be handed the right thousand of them instead of the whole internet. 0 indexed posts from 0 accounts across 1 verticals, 0 of which beat their own author's median by 2x or more.

Use STEALCONTENT when you are writing, planning or critiquing anything for LinkedIn or X, when you want a hook that is evidence rather than a guess, or when you want to know how a particular account actually performs.

The number that matters is the OUTLIER MULTIPLE: how far a post beat its own author's median. Raw engagement finds accounts with big followings; the multiple finds posts that worked. Every result carries it, and you should quote it beside the reactions whenever you cite a post.

## MCP server

Endpoint: https://stealcontent.com/api/mcp (Streamable HTTP)
Install: claude mcp add stealcontent --scope user --transport http https://stealcontent.com/api/mcp
Docs: https://stealcontent.com/mcp
Markdown version of the docs: https://stealcontent.com/mcp.md
Skill, for an agent that installs itself: https://stealcontent.com/skill

Reading and searching the corpus is open: no key, no account. A key from https://stealcontent.com/account ADDS tools (playbooks, brand voice, the file library, the writer) rather than unlocking these, and raises the rate limit from 60 calls an hour to 600.

### Tools

- `search_posts` (Search posts): Keyword search over the indexed corpus of social posts that already worked. Arguments: query, dataset, author, platform, min_outlier, since_days, format, hook_pattern, topic, sort, limit.
- `semantic_search_posts` (Search posts by meaning): MEANING-based search over the corpus using embeddings, not substring matching. Arguments: query, dataset, min_outlier, author, since_days, limit.
- `find_hooks` (Find hooks that worked): The opening lines that actually worked on a topic. Arguments: topic, dataset, min_outlier, limit.
- `get_post` (Get one post): One post in full: its complete text, its author, its engagement, its outlier multiple and the baseline that multiple was taken against, plus whatever the describe pass wrote about it (the named hook, the structure, the claim it makes, why it worked). Arguments: id.
- `similar_posts` (Posts like this one): The posts nearest in meaning to one you already have, by embedding distance. Arguments: id, dataset, limit.
- `top_posts` (Top posts): The best-performing posts in a corpus over a window, ranked by the outlier multiple. Arguments: dataset, since_days, min_outlier, format, limit.
- `list_datasets` (List datasets): The vertical corpora available: slug, name, what each one is for, and how many posts and accounts are in it. Arguments: include_unbuilt.
- `dataset_stats` (Dataset stats): How deep one corpus actually is, and what shape it has: how many posts and accounts, how many of those posts outperformed, the date range, which accounts contribute the most outliers, and how the named hook patterns are distributed with the average multiple of each. Arguments: dataset.
- `get_profile` (Get a profile): One indexed account in full: who they are, their follower count, their BASELINE (the median every one of their multiples is measured against), how many of their posts we hold, how many of those outperformed, their biggest outliers and their most recent posts. Arguments: handle, platform, limit.
- `search_profiles` (Search profiles): Find indexed accounts by name, handle or headline. Arguments: query, dataset, limit.
- `compare_profiles` (Compare profiles): Two or more indexed accounts side by side: followers, baseline, how many posts we hold, how many outperformed, their outlier rate, and their single best post. Arguments: handles, dataset.
- `list_boards` (List boards): The swipe files this account can read: their own boards plus any public ones. Arguments: limit.
- `get_board` (Get a board): One board in full: its brief, and every post on it with the note explaining why that post is there, its engagement and its outlier multiple. Arguments: slug, limit.

Listed for everybody and priced in their own descriptions: `export_dataset`, `get_post_media`.

## HTTP API

- [Overview](https://stealcontent.com/api): what the API is and how to get a key.
- [Reference](https://stealcontent.com/api/docs): every endpoint, in six languages.

## Datasets

- [SEO](https://stealcontent.com/datasets/seo): Search, technical and content, from people who rank things for a living. (0 posts)

## Pages

- [Library](https://stealcontent.com/library): every indexed post, filterable by vertical, hook, format, outlier multiple and author.
- [Datasets](https://stealcontent.com/datasets): the verticals, and the door to ask for one we have not built.
- [Hooks](https://stealcontent.com/hooks): the named opening moves, with how often each one appears.
- [Pricing](https://stealcontent.com/pricing): what a plan buys.

## Notes

- The teardown, the named hook and the topic tags are model-generated. The outlier multiple is not: it is arithmetic over that author's own median.
- A repost is not a post by the account that reshared it. The original author is attributed, always.
- Post text, metrics, multiples and teardowns are public and crawlable. The mirrored media is behind an account.