GitHub ProjectsSAFE
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
A Model Context Protocol (MCP) server that provides tools for managing GitHub Projects V2 through Claude and other MCP clients. This server uses the GitHub GraphQL API for interacting with GitHub Projects V2.
Features
- List and view GitHub Projects V2 for users and organizations
- Get project fields and items (issues, PRs, draft issues)
- Create issues and add them to projects
- Create draft issues directly in projects
- Update project item field values
- Delete items from projects
Usage
This server can be used with any MCP client, such as Claude Desktop. Add it to your MCP client configuration (e.g., claude_desktop_config.json).
Option 1: Using Published Package
Here's an example configuration using uvx as the command runner:
{
"mcpServers": {
"github-projects": {
"command": "uvx",
"args": [
"mcp-github-projects"
],
"env": {
"GITHUB_TOKEN": "your_pat_here"
}
}
}
}Make sure to replace your_pat_here with your actual GitHub Personal Access Token.
Option 2: From Source Code
To run the project directly from source code, follow these steps:
Setup
- Clone the repository:
git clone [email protected]:Arclio/github-projects-mcp.git cd github-projects-mcp
- Create and activate a virtual environment:
uv venv source venv/bin/activate
- Install dependencies:
uv pip install -e .
- Set your GitHub token as an environment variable:
Create a GitHub Personal Access Token and give it the necessary permissions. The required permissions are: repo, project, and read:org.
Add the token to your .env file after running this command:
cp .env.example .env
Then add the following to your .env file:
export GITHUB_TOKEN=your_personal_access_token
Usage from source code
468c188444fdOBSERVED · 2026-10-08Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add mcp-github-projects --env GITHUB_TOKEN=${GITHUB_TOKEN} -- uvx mcp-github-projects{
"mcpServers": {
"mcp-github-projects": {
"command": "uvx",
"args": [
"mcp-github-projects"
],
"env": {
"GITHUB_TOKEN": "${GITHUB_TOKEN}"
}
}
}
}Exposed tools (8)
3 read · 4 write · 1 destructive. Blast radius: 1 tool can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
add_issue_to_project | write | Add an existing GitHub issue to a Project V2. |
create_draft_issue | write | Create a draft issue directly in a GitHub Project V2. |
create_issue | write | Create a new GitHub issue. |
delete_project_item | destructive | Delete an item from a GitHub Project V2. |
get_project_fields | read | Get fields available in a GitHub Project V2, including options for SingleSelect fields. |
get_project_items | read | Get items in a GitHub Project V2. Can filter by state OR a single custom field=value. |
list_projects | read | List GitHub Projects V2 for a given organization or user. |
update_project_item_field | write | Update a field value for a project item. |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (2 observation(s))
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (1)
delete_project_item
Gates applied: no_behavioural_pass.
468c188444fdfull audit observations/trust-audit/mcp-server/arclio__github-projects-1.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-08 | 468c188444fd | SAFE | B | 89 | first audit |
Questions
What tools does GitHub Projects expose?
8 in total: 3 read-only, 4 that write, and 1 that can delete or overwrite (delete_project_item). Every one is listed on this page with its risk.
Is GitHub Projects safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 1 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does GitHub Projects need?
It reads GITHUB_TOKEN from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does GitHub Projects run?
It speaks stdio, so it runs as a local process your client starts. It is published on PyPI as mcp-github-projects.
How current is this page?
The grade is for one exact copy of the source (468c188444fd), read on 2026-10-08. The repository is watched and re-audited when it changes.