For a local default instance connect with `redis-cli` (defaults to 127.0.0.1:6379, db 0).
To specify host/port/db use `redis-cli -h <host> -p <port> -n <db>`. -n selects the logical database index (0-15 by default in standalone Redis).
Authenticate with `redis-cli -a <password>` or `--user <user> --pass <password>` (ACL). Prefer AUTH via the `REDISCLI_AUTH` env var or `redis-cli -a` interactive prompt rather than on the command line so the password isn't in shell history/ps.
For TLS use `redis-cli --tls` (optionally `--cacert`, `--cert`, `--key` for mutual TLS).
Test connectivity with `redis-cli ping` -> should return PONG. Check `redis-cli INFO server` for version.
Use a connection URL form with `redis-cli -u redis://[[user]:[pass]@]host:port[/db]` (scheme redis:// for plain, rediss:// for TLS).
Known gotchas
Default port 6379, db 0; -n picks a logical DB which is NOT a namespace for security - ACLs apply per-user, so don't rely on separate DBs for isolation.
Putting -a password on the command line exposes it via process list and shell history; use REDISCLI_AUTH instead.
Redis CLI stays in a connected REPL after a command - pass `--no-auth-warning` and use `--raw` if you need clean output for scripting.
Older Redis (pre-6) had no ACL users and used a single requirepass password.
Give your agent this knowledge — and 17,500+ more routes
One MCP install gives any agent live access to the full route map across 6,000+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?