Skip to main content

tinkr tail

Stream live logs to your terminal. Polls the backend continuously and prints new entries as they arrive.

tinkr tail <service> [options]

Options​

FlagShortDefaultDescription
--query TEXT-q*Filter expression (e.g. level:ERROR)
--lines INT-n—Show last N lines first, then stream live
--poll FLOAT-p2.0Poll interval in seconds
--resource TEXT-r—Resource type: ecs, lambda, cloudrun, gke, etc.
--output FORMAT-otableOutput format: table, json, jsonlines

Examples​

# Stream all logs
tinkr tail payments-api

# Show last 20 lines, then stream live (like tail -n 20 -f)
tinkr tail payments-api -n 20

# Errors only
tinkr tail payments-api -q 'level:ERROR'

# Last 50 error lines then stream
tinkr tail payments-api -n 50 -q 'level:ERROR'

# Pipe to jq
tinkr tail payments-api --output jsonlines | jq .message

Output​

Tailing payments-api · last 20 lines  (Ctrl-C to stop)

--- live stream ---
14:01:03  ERROR     Payment charge failed: card_declined
14:01:04  ERROR     Stripe API timeout after 30s
14:01:09  INFO      Health check OK
14:01:14  ERROR     Payment charge failed: insufficient_funds

When -n is used, the historical lines print first (oldest at the top), followed by a --- live stream --- divider, then live entries as they arrive.

Press Ctrl-C to stop.

Notes​

  • tinkr tail polls the backend — it is not a WebSocket push stream. Most observability backends don't expose a native push API.
  • The -n lookback window is 30 minutes. For older history use tinkr logs --since.
  • For high-volume services, use -q to reduce noise before piping to other tools.

See also​