Skip to security report content
← All security reports

MCP servers · Security report

Alerts to review

github/github-mcp-serverIs github-mcp-server safe to use?

Short answer

No scan can prove a project is safe. Our latest scan of github-mcp-server flagged 8 automated alerts, all classified as Secrets, and 1 of them are high or critical. Review them before you run github-mcp-server with real data or credentials.

Automated alerts
8
High or critical
1
Most flagged
Secrets
Last scanned

GitHub's official MCP Server

GoMIT33,627 GitHub stars

The assessment, at a glance

All 8 automated alerts are classified as Secrets

100% of the alerts in the current repository snapshot are classified as Secrets. 1 of all alerts are critical or high severity. Categories and severity are triage signals; these alerts have not been independently validated as vulnerabilities.

Repository snapshot from Plexicus staging. Counts may include earlier scans and are not attributed to an exact run or commit. Check categories come from the latest observed job and can differ from the scans contributing findings. Completed categories describe bundle delivery, not full repository coverage.

What we found

What the scan flagged

Alerts are leads for review, sorted by type and severity. They are not confirmed vulnerabilities.

Alerts by type

Secrets100% of alerts
8

Types describe all current alerts for this repository, not only the latest scan.

Repository findings snapshot

Current repository findings

8automated alerts

Findings currently visible in the Plexicus repository record. These alerts have not been validated as vulnerabilities or attributed to this exact scan or commit.

Critical
0
High
1
Medium
7
Low
0
Informational
0

Source: Plexicus staging · Collected

Scope and counting limits

These are current repository totals, which may include findings from earlier scans. Scan-specific counts and the scanned commit remain unverified. The snapshot excludes findings marked as false positives, duplicates, deleted, or sandbox results, and findings still awaiting processing or enrichment.

Next steps

How to use github-mcp-server more safely

Practical steps based on what the scan flagged. They are general guidance, not confirmation of an exploitable vulnerability.

  1. Review the 1 critical or high severity alerts first; confirm validity and impact before remediation.

  2. Confirm false positives and duplicate root causes before using alert totals to prioritize work.

  3. Secrets (8): validate suspected secrets privately. Revoke or rotate credentials only when exposure is confirmed, and never publish secret values.

  4. Try it in a container or virtual machine first, with throwaway credentials.

  5. Pin a specific release or commit instead of tracking the latest code.

This public report shows totals only. File names, exploitability, and false-positive rates are not included.

What we checked

Which checks ran?

Each card shows whether the latest scan completed that type of check. A completed check does not mean every file or rule was covered.

Secrets

Checked

The latest scan completed this check. It does not guarantee full coverage of the repository.

Code security

Not checked

Do not assume this area was assessed.

Dependencies

Not checked

Do not assume this area was assessed.

Configuration

Not checked

Do not assume this area was assessed.

Limits

What this report cannot tell you

Technical scan record & repository metadata
Platform execution
Platform job completed
Commit evidence
Scanned commit not verified
Last checked (UTC)
2026-10-16T05:03:37.349Z
Default branch
main

Result bundles: 1 of 1 complete; 0 failed; 0 degraded. Bundle completion alone does not establish full scan coverage.

Requested commit e3809fe91c7546633241006bf0b91f23b2595e28

Scan source fork: View fork ↗

Repository metadata retrieved 2026-10-15T08:13:45.762Z. Candidate commit from metadata: c59e8c199efbf14467211861de44afaeeb6873e3. A candidate commit is not proof of the scanned commit.

Questions about github-mcp-server

Is github-mcp-server safe to use?

No scan can prove a project is safe. Our latest scan of github-mcp-server flagged 8 automated alerts, all classified as Secrets, and 1 of them are high or critical. Review them before you run github-mcp-server with real data or credentials.

What did Plexicus find in github-mcp-server?

100% of the alerts in the current repository snapshot are classified as Secrets. 1 of all alerts are critical or high severity. Categories and severity are triage signals; these alerts have not been independently validated as vulnerabilities.

What did the scan not check?

Runtime behavior, deployment settings, MCP and tool permissions, and prompt injection are not covered unless listed in the checks above. Review those before giving the tool access to real data.

How can I check my own fork or project?

Scan it with the free Plexicus SAST tool. Connect GitHub and get a report for your own repository.

For maintainers

Add the github-mcp-server security report to your README

Show users where to find this report. The badge links to this page and makes no safety claim.

Plexicus security report
[![Plexicus security report](https://www.plexicus.ai/open-source/badge/github/github-mcp-server.svg)](https://www.plexicus.ai/open-source/github/github-mcp-server/)

Your own code

Scan your own repository for free.

Connect GitHub and see what Plexicus flags in your code.

Scan your repository free ↗
Qualification

Check whether AI Swarm Pentest fits your environment.

Share the minimum context. We will review the scope and tell you the next commercial step.

Before submitting — verify you fit

0 / 280

No commitment. If you don't fit, we'll tell you.

SAMPLE HANDOVER · ILLUSTRATIVE

Sample evidence handover

A trimmed view of what your team receives at the end of an AI Swarm Pentest engagement. Real engagements include full technical evidence, executive narrative, and a remediation plan.

VALIDATED FINDING Evidence attached

Server-Side Request Forgery in webhooks/receiver

demo-project/sample-app · src/webhooks/receiver.py:42

SeverityHigh CVSS 3.18.6 Priority79 Confirmedvia replay

Untrusted caller-supplied URLs reach an internal egress without an allowlist. Replayed in a sandbox against a fresh authorized target — the same control was validated to fail twice.

REVIEWER-READY REMEDIATION Merge-ready PR

Validate the target URL against an allowlist of permitted hostnames. Reject private/internal IP ranges. Enforce HTTPS only.

plexicus/remediation/webhooks-ssrf 3 changed · 0 new files
42resp = requests.get(target_url)
42+if not is_allowed_host(target_url):
43+  raise WebhookRejected(target_url)
44+resp = requests.get(target_url, timeout=5)
Every engagement hands over:
  • Executive briefing
  • Validated findings list
  • Merge-ready PRs
  • Compliance mapping (NIS2 · DORA · CRA)
Private Round For investors