Skip to main content
Glama

rf-log-mcp

An MCP server for inspecting Robot Framework result files, providing LLMs with concise evidence views.

Feature Overview

Supported inputs:

  • output.xml: Robot / Rebot 6.0.x / 6.1+ / 7.x

  • output.json: Robot / Rebot 7.2+

Exposed MCP capabilities:

  • Tools

    • parse_result

    • get_view

    • search_messages

  • Resources

    • rf://runs/{run_id}/summary

    • rf://runs/{run_id}/tests/{test_id}

Supported views:

  • summary

  • failure_path

  • step_window

Related MCP server: RobotMCP

Key Notes

  • This project is an MCP stdio server

  • The correct usage is: The MCP host starts the rf-log-mcp process and then calls tools and resources via stdio


Quick Start

1. Install dependencies

uv sync
{
  "mcpServers": {
    "rf-log-mcp": {
      "command": "uv",
      "args": [
        "run",
        "python",
        "-m",
        "rf_log_mcp"
      ]
    }
  }
}

Packaging and Installation

Build

uv build

Generated after build:

  • dist/rf_log_mcp-0.1.0-py3-none-any.whl

  • dist/rf_log_mcp-0.1.0.tar.gz

Install wheel

uv pip install dist/rf_log_mcp-0.1.0-py3-none-any.whl

After installation, you can start it directly:

rf-log-mcp

MCP configuration example for installed package

{
  "mcpServers": {
    "rf-log-mcp": {
      "command": "rf-log-mcp",
      "args": []
    }
  }
}

Windows explicit path example

{
  "mcpServers": {
    "rf-log-mcp": {
      "command": "D:\\project\\rf_log_mcp\\.venv\\Scripts\\rf-log-mcp.exe",
      "args": []
    }
  }
}

Typical Call Flow

Step 1: Parse result file

parse_result(path="tests/fixtures/single_failure_611.xml")

Typical return:

{
  "ok": true,
  "run_id": 1,
  "source_format": "xml"
}

Step 2: Get summary

get_view(run_id=1, view="summary")

Step 3: Get failure path or search messages

get_view(run_id=1, view="failure_path")
search_messages(run_id=1, query="timeout")

Environment Variables

RF_LOG_MCP_DB

Used to override the default SQLite database path.

PowerShell example:

$env:RF_LOG_MCP_DB="D:\data\rf-log-mcp\store.sqlite3"
rf-log-mcp

MCP configuration example:

{
  "mcpServers": {
    "rf-log-mcp": {
      "command": "rf-log-mcp",
      "args": [],
      "env": {
        "RF_LOG_MCP_DB": "D:\\data\\rf-log-mcp\\store.sqlite3"
      }
    }
  }
}

FAQ

1. Why use uv?

  1. Local environment dependencies may conflict with MCP, requiring venv to isolate dependency conflicts (conda is too slow, uv is fast)

2. Can get_view / search_messages accept file paths?

Yes. If the file has already been parsed, the service will convert the path to the corresponding run_id before querying. However, it is still recommended to prioritize using the integer run_id returned by parse_result().

5. Under what circumstances can this project not be used directly?

If your LLM platform:

  • Does not support MCP

  • Or does not support starting local processes

Then it cannot be integrated directly and requires an additional integration layer.


Development Checks

uv run ruff check .
uv run pytest

Available Tools

3 tools
get_viewC

Get one supported view using a numeric run id or an already-parsed file path.

Inputs: run_id (required), view (required), selector, cursor, +1 more. Returns: object

parse_resultB

Parse a Robot Framework output.xml or RF 7.2+ output.json file and index it.

Inputs: path (required), force_rebuild. Returns: object

search_messagesC

Search indexed messages using a numeric run id or an already-parsed file path.

Inputs: run_id (required), query (required), level, limit, +1 more. Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.3/5.0

    Scored across 3 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose: get_view retrieves a specific view, parse_result processes and indexes files, and search_messages queries indexed messages. There is no overlap in functionality, and the descriptions clearly differentiate their roles, making misselection unlikely.

    Naming Consistency5/5

    All tool names follow a consistent verb_noun pattern (get_view, parse_result, search_messages) with clear, descriptive verbs and nouns. The naming is uniform throughout, using snake_case consistently without any deviations or mixed conventions.

    Tool Count3/5

    With only 3 tools, the set feels thin for a Robot Framework log management server, as it might lack operations like updating or deleting parsed data. However, the tools cover core parsing and querying functions, making it borderline appropriate but potentially limited in scope.

    Completeness3/5

    The tools provide basic parsing and search capabilities, but there are notable gaps: no update or delete operations for indexed data, and no tools for managing multiple runs or views beyond retrieval. This could lead to dead ends in workflows requiring data modification or broader management.

    Maintenance

    ActivityNo data
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      MCP server for log file analysis. Gives LLMs the ability to efficiently analyze large log files without loading them into context.
      7
      101
      MIT
    • A
      license
      B
      quality
      B
      maintenance
      RobotMCP is a comprehensive Model Context Protocol (MCP) server that bridges the gap between human language and Robot Framework automation. It enables AI agents to understand test intentions, execute steps interactively, and generate complete test suites from successful executions.
      19
      4,288 PyPI
      118
      Apache 2.0
    • A
      license
      Not graded
      quality
      D
      maintenance
      MCP server that helps resolve Robocop static code analysis errors and warnings via LLM, providing tools to get reports and run format.
      11
      MIT