Configure LinkBunny in Cursor's editor and CLI

Install the shared LinkBunny Skill

Review the canonical Skill, then use the setup prompt or the manual command below. The Skill holds the API workflow and approval rules. Local MCP is optional; both interfaces follow the same instructions.

In bash or zsh, this downloads the Skill to ~/.agents/skills/linkbunny/SKILL.md and refuses to overwrite a copy. It needs curl with --remove-on-error support. For one repository, save the reviewed file at .agents/skills/linkbunny/SKILL.md instead. Keep credentials out of the Skill.

skill_dir="$HOME/.agents/skills/linkbunny"
skill_file="$skill_dir/SKILL.md"
mkdir -p "$skill_dir" &&
if [ -e "$skill_file" ]; then
  printf 'Refusing to overwrite %s\n' "$skill_file" >&2; exit 1
else
  curl -fsSL --remove-on-error "https://getlinkbunny.com/skill.md" -o "$skill_file"
fi

Cursor reads .agents/skills and .cursor/skills at project and user scope. It also loads Claude- and Codex-compatible skill directories. If you installed LinkBunny for another agent, check for that copy before installing again. In a new local chat, ask Cursor to find linkbunny, show its loaded file path and read it fully.

A personal Skill stays on this machine. It does not give a Cloud Agent or remote SSH session that file, your token or your publication access. Use this guide for the local editor and local agent CLI.

Subscribe before creating a marketplace token

$20/month covers up to 10 active sites. Sign in to Account, subscribe, then create a token. Approve account, subscription and credential actions as they arise. Your agent's account and model usage are separate from LinkBunny's plan.

Store LINKBUNNY_API_TOKEN in an approved secret store or a private, Git-ignored file. The local MCP examples use --token-file: only the absolute file path goes in configuration. The loader reads that variable without executing the file as a shell script. Keep the value out of prompts, command arguments, source, URLs, logs and screenshots.

Have the agent propose a profile from your site's content before registration. Each new profile needs manual marketplace review before matching. Approve one check-in at a time; check_in can reserve an assignment. Review each contextual nofollow edit, approve publication, then approve reporting separately. Approve rejection when no suitable context exists. Stop when review is pending or no request is available; do not poll automatically.

The agent workflow owns these steps. The directory backlink tutorial uses the free hosted checker for an existing link and needs no LinkBunny signup or token.

Add one project entry to .cursor/mcp.json

The local MCP process needs Node.js 22+ and npm/npx on the Cursor host. Have Cursor prepare the config diff, pinned package, secret-file path and trust steps. Review that bundle before approving the edit. Preserve other servers and check ~/.cursor/mcp.json for an existing LinkBunny entry.

Prepare project-scoped registration for the local stdio server @linkbunny/mcp@0.1.1 in this coding agent. Follow https://getlinkbunny.com/docs/agents#mcp for the correct client configuration, secret loading, trust, restart, and safe verification. Never reveal or copy the LinkBunny API token into chat, command arguments, logs, screenshots, or tracked configuration. Before changing persistent MCP configuration, show me the exact client, scope, config diff, pinned command, secret-loading path, and restart/trust steps, then ask for approval. After setup, read linkbunny://guidance/skill before using any LinkBunny tool. Do not call a LinkBunny tool merely to test setup.

{
  "mcpServers": {
    "linkbunny": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@linkbunny/mcp@0.1.1",
        "--token-file",
        "/absolute/project/.env.linkbunny"
      ]
    }
  }
}

Replace the example absolute path with the private file on this host. Cursor's root key is mcpServers; command and args are separate. Cursor also supports envFile for stdio servers, but that loads variables from the file into the process. This example uses LinkBunny's loader to read only LINKBUNNY_API_TOKEN.

Cursor configuration and Skill directories · checked 15 Oct 2026: MCP · Skills

Keep server trust and action approval separate

Approving an MCP server lets Cursor connect to it. It does not approve a check-in, publication or report. In the editor, use the Allowlist Run Mode with an empty allowlist and sandboxing disabled when shell and marketplace calls must pause for review. With sandboxing enabled, supported shell commands can still run automatically. Approved shell commands then run outside Cursor's sandbox, so inspect each command carefully.

Auto-review can approve calls through a classifier, so leaving LinkBunny off a tool allowlist in that mode is not enough. Keep marketplace tools and publication commands out of automatic approvals, and retain the Skill's approval before each action.

In the CLI, review the active global/project permissions and avoid persistent Mcp(linkbunny:*) approvals or blanket tool-approval flags. Follow the Skill's per-action approval rules in the conversation as well. Cursor Cloud Agents run without these local Run Mode pauses; keep marketplace tools and credentials in the interactive local workflow.

Cursor editor and CLI approval behavior · checked 15 Oct 2026: Run Modes · CLI permissions

Inspect the server without starting a link request

agent mcp list
agent mcp list-tools linkbunny

Review the configuration source and tool declarations. If Cursor asks you to approve the server, inspect the command and path first. Its CLI also provides agent mcp enable linkbunny to add that server to the local approved list; use it only after your setup review. Keep action approvals enabled afterward.

The pinned local package exposes register_profile, update_profile, check_in, report_assignment and reject_assignment. Read the installed Skill before use. Inspecting tools does not require calling one, and check_in can reserve an assignment. The hosted checker's single inspect_placement tool belongs to a different server.

For a Cursor authentication error, follow Cursor's account login flow. Cursor documents agent login and API-key authentication; LinkBunny's free checker does not remove Cursor's own account or model requirements. Keep both services' credentials private.

Cursor CLI commands and authentication · checked 15 Oct 2026: CLI MCP · CLI authentication

Resolve duplicate Skills, scope and approvals

Cursor reads an old Skill

Inspect copies in .agents, .cursor, .claude and .codex, including personal directories. Keep one reviewed copy and start a new session. Updating the MCP package does not replace an installed Skill.

The server is awaiting approval

Review the native server-trust prompt, then inspect tools again after approval. Don't use a marketplace operation as a test. If the server is disconnected, check Node.js, npx and the token-file path on the host running Cursor.

Actions run without pausing

Inspect Run Mode, tool allowlists, CLI permissions and team policy. Auto-review and an empty tool allowlist are different controls. Stop the workflow until you can review each action; don't move the token into a Cloud Agent to bypass local approval.

Compare other coding-agent setup paths.