LinkIntel
Evidence for your AI's next X post.
Community: Submitted by a user or imported; check the owner before granting accessOnlineNo sign-inGlobalFreeCan modify data
What it can do
- Get Data Status: Report LinkIntel readiness per capability: metric freshness, registered and measured post counts, connected corpus sources, owned-X connection status, LinkedIn connection status, rema
- Prepare Content Brief: Prepare a stored-data writing brief for a topic and an explicitly named public X reference cohort. Costs nothing and never fetches: it reads references already stored by add_ref
- List References: List the stored reference topics and cohorts for this account, with which platform and how many posts each holds. Call this before prepare_content_brief so the topic and cohort are th
What data it sees
Do you need an account
No: the server works without sign-in
Evidence for your AI's next X post. Connect X and LinkIntel reads your recent original posts and their metrics daily, tracks X's open-source ranking weights, and stores public X posts you pick as references for a topic. Your assistant gets it all through prepare_content_brief as one cited content brief, with the limits of the evidence stated plainly. It never publishes, never schedules and never searches X on its own. OAuth sign-in, $39/month.
How to connect
How to connect LinkIntel to Claude
The server address to paste into Claude settings:
https://linkintel--marcos.run.tools
- Open Claude (claude.ai or the desktop app).
- Go to Settings → Connectors.
- Click Add custom connector.
- Paste the server address copied below into Remote MCP server URL and click Add.
- In a chat, click + → Connectors and switch the new connector on.
Custom connectors are available on Free, Pro, Max, Team and Enterprise plans (Free is limited to one). On Team and Enterprise an organization Owner adds the connector first under Organization settings → Connectors.
Authorization
After you click Add, a sign-in window for the service opens. Sign in with your own account and approve access. Claude never sees your password.
How to connect LinkIntel to ChatGPT
The server address to paste into ChatGPT settings:
https://linkintel--marcos.run.tools
- Open ChatGPT in a browser (chatgpt.com). A Plus, Pro, Business, Enterprise or Edu plan is required.
- Turn on developer mode once: Settings → Apps → Advanced settings → Developer mode.
- Open Settings → Connectors and click Create.
- Fill in the form: Name (anything), Description (one line about what the service does), MCP server URL (copy it below).
- Under Authentication choose OAuth if the service requires sign-in, otherwise None. Click Create.
- In a new chat open + → Apps/Connectors and enable the connector.
OpenAI has renamed this section before (Connectors → Apps/Plugins). If the label differs, search settings for "developer mode". On Business/Enterprise workspaces an admin must allow custom connectors first.
Authorization
On first use ChatGPT opens the service's sign-in window. Sign in and approve access.
How to connect LinkIntel to Cursor
The server address to paste into Cursor settings:
https://linkintel--marcos.run.tools
Fastest: click Open in Cursor below and confirm the prompt.
Manually:
- In Cursor open Settings → Cursor Settings → MCP and click Add new global MCP server.
- Paste the JSON copied below into
~/.cursor/mcp.json(per project:.cursor/mcp.jsonin the repo root). - Save the file. The server appears in the MCP list; authorize it there if asked.
Authorization
If the service needs sign-in, an authorize button appears next to the server in the MCP list.
Server tool list (8)
Raw names from tools/list. Only developers need these.
| get_data_status | Report LinkIntel readiness per capability: metric freshness, registered and measured post counts, connected corpus sources, owned-X connection status, LinkedIn connection status, remaining reference lookups, token expiry and today's API budget. The top-level connected field reports MCP connection status. Use linkedin_connected for LinkedIn readiness and the ready object to check specific capabilities. Call this before the first analysis and whenever a number looks wrong or surprisingly low. |
| prepare_content_brief | Prepare a stored-data writing brief for a topic and an explicitly named public X reference cohort. Costs nothing and never fetches: it reads references already stored by add_references, so call list_references first to use the exact stored labels. platform picks whose owned evidence the brief reads, because a post's length and structure do not transfer between platforms: "x" (the default) reads the creator's most recent stored X posts, which are not filtered to the topic, so the response asks the agent to read the excerpts and judge relevance itself; "linkedin" reads the creator's historical LinkedIn posts already tagged with this topic. The reference cohort is always public X posts, and the response says so plainly when platform is "linkedin" so that evidence is never mistaken for a same-platform comparison. Neither a LinkedIn connection nor a live platform fetch is required for cached references. Never searches X or predicts reach. |
| list_references | List the stored reference topics and cohorts for this account, with which platform and how many posts each holds. Call this before prepare_content_brief so the topic and cohort are the exact stored labels rather than a guess, before add_references so a near-duplicate label is not created, and before remove_references (with include_posts) to see exactly which posts a removal would affect. Reads stored data only and costs nothing. |
| find_account_posts | Look up a public X account's most recent original posts (replies and reposts excluded) and return their URLs, dates and short previews for the user to review. This draws on the same daily read budget add_references uses, and stores nothing: it never writes a reference, so it can be called freely to explore before anything is added. Only the "x" platform is supported today. |
| add_references | Store public reference posts as evidence under a topic and a named comparison cohort, so prepare_content_brief can use them. Only the "x" platform is supported for reference import today. Reads only the URLs supplied. It never searches the platform, connects an account, collects the user's own analytics, or publishes anything. |
| remove_references | Remove stored reference posts: specific posts by URL, or a whole topic and cohort group. Removing shapes every brief prepare_content_brief builds from what remains, so confirm the exact posts or group with the user first; call list_references with include_posts to show exactly what is stored before removing any of it. Prefer removing specific posts over a whole group where the user only meant to drop a few. Supply urls (optionally narrowed by topic and/or cohort) to remove those specific posts, or topic and cohort together with no urls to remove a whole group. It never searches or reaches any platform: it only deletes rows already stored by add_references. |
| get_owned_performance | The tool for questions about the user's own X posts: 'how did my posts do', 'what was my best post', 'which posts got the most engagement'. Returns, per post, impressions, likes, replies, reposts, quotes and bookmarks, plus the owner's non-public and organic metrics where available, read from stored, profile-scoped snapshots of the connected account. Covers recent original posts only, not replies or reposts. Use sort_by to rank the returned posts by recency, impressions, total engagement, or engagement rate. It never searches the platform, imports posts, makes a live platform request, or predicts reach. |
| get_platform_signals | Return X's published default ranking weights (what helps and what hurts, each with the commit link), the material weight changes from the last 14 days, and how to use them when drafting or reviewing a post. Reads stored data only and costs nothing: it never fetches X or GitHub live, and the values are defaults mirrored from X's open-source repo, synced daily. It describes what X's published ranking defaults reward in aggregate, not a reach prediction for any post. Use the signals as a final check that never overrides the evidence or the creator's voice. Only "x" has a readable source; other platforms return available: false. |