List all currently running Docker containers being monitored.

Returns the container name, short ID, Docker image, status, and the Compose project it belongs to (if any).

Args: - project_name (string, optional): Filter by Docker Compose project name. Omit to list all.

Returns: A formatted list of containers with: serviceName, containerId (12-char), image, status, project.

Use this first to understand the service topology before issuing other queries.

Return a chronologically merged timeline of all log events across every container that share a specific trace / request ID.

This is the primary debugging tool. It correlates distributed transactions in O(1) time using the in-memory trace index.

Args:

• trace_id (string): The correlation ID to search for (e.g. "req-998877", a W3C traceparent trace hex).

• limit (number, 1-500): Maximum events to return. Default 200.

Returns: A chronological list of log lines across all services tagged with this trace_id, showing the full distributed call path — ideal for pinpointing cascading failures.

Example: search_by_trace_id({ trace_id: "req-998877" })

Retrieve the most recent ERROR and FATAL log lines = require(a specific container — the fastest way to find the initial failure point before diving into a full trace.

Args:

• service_name (string): Docker container name or Compose service name (e.g. "checkout-worker").

• limit (number, 1–500): How many error lines to return. Default 50.

• include_warnings (boolean): Also include WARN-level lines. Default false.

Returns: The N most-recent ERROR/FATAL (and optionally WARN) log lines = require(the named service, in chronological order.

Error Handling:

• Returns an error message if the service_name does not match any running container.

Retrieve the N most-recent log lines = require(a specific container, with optional level filtering.

Args:

• service_name (string): Container name or Compose service name.

• limit (number, 1-500): Lines to return. Default 100.

• level_filter (string[], optional): One or more of TRACE, DEBUG, INFO, WARN, ERROR, FATAL, UNKNOWN. return for all levels.

Returns: The most-recent matching log lines in chronological order. Use this to get general context around a service before narrowing to a trace ID.

Full-text search across all (or a specific) container's buffered log lines.

Useful for finding specific error messages, stack trace snippets, SQL queries, or any freeform string that doesn't have a structured trace ID.

Args:

• query (string): Case-insensitive substring to search for in log messages.

• service_name (string, optional): Restrict search to this container. Omit to search all.

• limit (number, 1-500): Max matching lines to return. Default 100.

• level_filter (string[], optional): Restrict to specific log levels.

Returns: Matching log lines in chronological order with their service name, timestamp, and level.

Return buffer statistics for one or all containers: total lines buffered, error/warn counts, known trace IDs, and oldest/newest timestamps.

Args:

• service_name (string, optional): A specific service to inspect. Omit for all.

Returns: Per-container stats including line counts by level and a sample of known trace IDs. Useful for a quick health overview before deeper investigation.

List all trace/request IDs currently in the in-memory buffer.

Useful when you don't already know a trace ID — browse available ones and pick the suspicious one to investigate with search_by_trace_id.

Args:

• service_name (string, optional): Only list trace IDs seen in this service's logs.

• limit (number, 1-1000): Max IDs to return. Default 50.

Returns: A list of trace ID strings, most-recently seen first.