Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@openbot-mcp check the server status"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
openbot-mcp is the Model Context Protocol server for
OpenBot.ai. It lets AI agents such as Claude Code, Cursor and Claude Desktop
use OpenBot tools for physical AI data work.
Status: skeleton. PyPI has only the pre-release
0.1.0.dev0, published to reserve the name;pip install openbot-mcpdoes not select it. Only theserver_statustool exists. Data audit, training readiness, Catalog and judgment tools are being designed.
pip install -e ".[dev]"
OPENBOT_MCP_ROOTS=/path/to/data openbot-mcpClaude Code:
claude mcp add openbot -e OPENBOT_MCP_ROOTS=/path/to/data -- openbot-mcpOther MCP clients use the same command in their MCP configuration file:
{
"mcpServers": {
"openbot": {
"command": "openbot-mcp",
"env": { "OPENBOT_MCP_ROOTS": "/path/to/data" }
}
}
}Related MCP server: MCP Workspace Server
Variable | Required | Description |
| Yes, for local tools | Directories the server may read, separated by the OS path separator. Paths outside them, including symlinks that point outside, are rejected. |
| No | Where artifacts are written. Default |
| No | OpenBot API base URL. Default |
| No | Default |
| No | Default |
Local tools only read the directories in OPENBOT_MCP_ROOTS and only write to the output
directory. The server never modifies source data, never runs shell commands and collects no
telemetry.
pip install -e ".[dev]"
python scripts/check_version.py
pytest -v
ruff check src tests scripts
mypy src
python -m buildVERSION is the package version source of truth. To release, update VERSION and
CHANGELOG.md, verify locally, then publish a GitHub Release whose tag is v<version>. The
release workflow tests every supported Python, builds the distributions and publishes to PyPI
through a trusted publisher.
openbot-mcp: MCP server for AI agents.
openbot-data: local robot/ego data processing library.
openbot-sdk: OpenBot platform API client.
MIT
Report the OpenBot MCP server version and which local directories it may read.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, so the safety profile is covered. The description adds a useful scoping detail ('which local directories it may read'), but it does not disclose any other behavioral aspects like potential network calls, authentication requirements, or response format specifics. Given the low bar set by annotations, this is adequate but not outstanding.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, front-loaded sentence with no filler. Every word contributes to explaining the tool's function. It is maximally concise while remaining informative.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The tool is simple with no parameters and an output schema, so return values are already covered. The description covers the essential information an agent needs to decide to call it and what to expect. Nothing critical is missing for a status-reporting tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has zero parameters, and schema coverage is 100% (vacuously). Per the rubric, with no parameters the baseline is 4. The description does not need to add parameter details, and it doesn't, so a 4 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a specific verb ('Report') and resource ('OpenBot MCP server version' and 'local directories it may read'). It is precise and unambiguous, and there are no sibling tools to differentiate from, so it fully establishes what the tool does.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies when to use it: whenever you need server version or directory access info. Since there are no alternatives or siblings, it doesn't need to explicitly route away from other tools. The clarity of purpose makes the usage obvious, though it doesn't spell out any conditions or exclusions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0server_statusScored across 1 tool
Only a single tool exists, so there is no possibility of confusion between tools. The tool's purpose is clear and distinct.
With only one tool, naming is trivially consistent. The name follows a clean noun_noun pattern with no conflicting conventions.
A single trivial status tool is far too few for an MCP server that presumably should provide meaningful bot-related functionality. This is an extreme mismatch.
The server only exposes a status/version endpoint and does not provide any actual operations, data access, or workflow coverage. The surface is severely incomplete for any plausible domain.
Operate Linux, macOS and Windows from your LLM. Every action runs through an auditable allowlist.
Safe folder access for ChatGPT and Claude: read, write and search files, risky tools opt-in.
Runtime permission, approval, and audit layer for AI agent tool execution.
Nifty's MCP server — exposes tasks, projects, messages, and files as tools for AI agents.
