l4 mcp status

l4 mcp status reports, per client, whether it is installed on this machine and whether it carries a LevelFour entry. Run this first when an agent says it cannot see any LevelFour tools.

l4 mcp status [flags]

Flags

Global flags (--json, --jq, --template, --quiet, --token, --api) apply. See CLI overview.

Output

LevelFour MCP
  Entry name: levelfour
  Hosted endpoint: https://mcp.levelfour.ai/mcp
  Local surface: 16 tools, 3 prompts
  Credential: system keychain (l4_live_...abcd)

╭────────────────┬───────────┬────────────┬──────────────────────────────────╮
│ Client         │ Installed │ Configured │ Endpoint                         │
├────────────────┼───────────┼────────────┼──────────────────────────────────┤
│ Claude Code    │ yes       │ yes        │ configured in Claude Code        │
│ Claude Desktop │ yes       │ yes        │ /opt/homebrew/bin/l4 mcp serve   │
│ Cursor         │ yes       │ no         │                                  │
╰────────────────┴───────────┴────────────┴──────────────────────────────────╯

Installed means the client was found on this machine. Configured means it carries an entry under --name. The Endpoint column shows what that entry actually points at, read back from the file, which is how you catch an entry left pointing somewhere unexpected.

If a client is installed but not configured, run l4 mcp install --client <id>.

Reading the credential line

Credential names where the key came from, in the order l4 resolves it: the --token flag, then LEVELFOUR_TOKEN, then the system keychain. If it reads none, run l4 auth login.

The key is masked. What matters is the source: a client configured with a key from your environment behaves differently from one configured with the key in your keychain, and this line tells you which you have.

Related

• l4 mcp install - configure a client that reports no
• l4 mcp uninstall - remove an entry it reports
• CLI authentication - how credentials resolve