MCP, MCP server
Mcp Server Gateway
Mcp Server Gateway, an MCP server listed in the PlugVet directory.
Compatibility: Not stated.
Grade D
Two or more Medium findings, or a High finding. Only try it in a sandbox.
Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.
Rescanned at commit f9698059ef79, unchanged since the check below.
PERMISSION_SHELL_UNDECLARED INSTALL_REMOTE_SCRIPT SECRETS_PATH_CREDENTIAL
Why this grade
- PERMISSION_SHELL_UNDECLARED Medium
Runs shell commands that the manifest or description does not mention.
Agents/jfrog-agent/scripts/create_bq_tracking_tables.py, line 64
subprocess.run(cmd, check=True, capture_output=True, text=True) - INSTALL_REMOTE_SCRIPT Medium
Fetches a script from the network and runs it (a pipe to a shell, or an install hook with a URL).
artifactory/.circleci/config.yml, line 183
curl -s https://fluxcd.io/install.sh | bash - SECRETS_PATH_CREDENTIAL Medium
Mentions a credential store path (~/.ssh, id_rsa, .npmrc, .aws/credentials, credentials.json, .netrc).
agent-safety-kit/agent_safety/input_guard/secrets.py, line 9
- Uploaded / pasted secret filenames (*.pem, credentials.json, …)
Any High, or two or more Medium findings, give D. One Medium gives C. Low findings or a security Note give B. Each code names a finding behind the letter.
If you think this grade is wrong, email [email protected] with the entry link and your evidence. We aim to acknowledge within two working days and decide within seven. If we confirm an error, we aim to correct it within 24 hours of confirming it. These are aims, not guarantees.
Grades are a hygiene check of the source at the commit shown, on the date checked, not a security audit or a guarantee. The free directory grades are automated. Before release, a PlugVet engineer reviews every paid report and any grade we hold back, including all F grades, against the source code and the evidence we quote.
What was checked
Plain notes from the hygiene checks.
Permissions
Concern
requests shell access. No manifest field names shell or command execution.
PERMISSION_SHELL_UNDECLARED
Agents/jfrog-agent/scripts/create_bq_tracking_tables.py, line 64
subprocess.run(cmd, check=True, capture_output=True, text=True)
Agents/jfrog-agent/scripts/create_bq_tracking_tables.py, line 90
acct = subprocess.run(["gcloud", "config", "get-value", "account"], capture_output=True, text=True, check=True)
Install and update
Serious
install fetches a script from the network.
INSTALL_REMOTE_SCRIPT
artifactory/.circleci/config.yml, line 183
curl -s https://fluxcd.io/install.sh | bash
artifactory/.github/workflows/build-and-deploy.yml, line 146
curl -s https://fluxcd.io/install.sh | sudo bash
Network
Concern
can reach a host that is not written as a fixed address.
NETWORK_NO_FIXED_HOST
agent-safety-kit/agent_safety/pii_scanner/alerter.py, line 76
with urllib.request.urlopen(req, timeout=timeout) as resp:
mcp-gateway/monitoring/grafana/build-dashboard.py, line 319
with urllib.request.urlopen(GRAFANA_UID_URL, timeout=3) as response:
mcp-gateway/scripts/e2e-test.mjs, line 18
const response = await fetch(`${GATEWAY}/${NAMESPACE}/mcp`, {
Show more
mcp-gateway/scripts/e2e-test.mjs, line 45
const health = await fetch(`${GATEWAY}/health`);
mcp-gateway/scripts/generate-rejections.sh, line 30
code=$(curl -s -o /dev/null -w '%{http_code}' -X POST "$BASE_URL$path" \
mcp-gateway/scripts/https-proxy.mjs, line 45
const proxyReq = http.request(
mcp-gateway/scripts/oauth-client-demo.py, line 86
resp = urllib.request.urlopen(req, context=_SSL_CTX)
mcp-gateway/src/client-handlers/cimd.ts, line 97
* Validates that the client_id URL is safe to fetch (HTTPS, not SSRF risk,
Agents/jfrog-agent/Dockerfile, line 39
CMD python -c "import urllib.request; urllib.request.urlopen('http://localhost:8501/_stcore/health')" || exit 1
Secrets
Concern
reads a credential or environment file.
SECRETS_ENV_READ SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV
Agents/jfrog-agent/jfrog_agent/llm.py, line 115
if not os.getenv("OPENAI_API_KEY"):
Agents/jfrog-agent/jfrog_agent/llm.py, line 129
if not os.getenv("ANTHROPIC_API_KEY"):
agent-safety-kit/agent_safety/input_guard/secrets.py, line 9
- Uploaded / pasted secret filenames (*.pem, credentials.json, …)
Show more
Agents/jfrog-agent/jfrog_agent/settings.py, line 46
os.getenv("JFROG_AGENT_TOKEN_CACHE", "~/.jfrog-agent/token.json")
Agents/jfrog-agent/jfrog_agent/settings.py, line 121
ctx_opt_max_tokens: int = int(os.getenv("CTX_OPT_MAX_TOKENS", "8000") or "8000")
Agents/jfrog-agent/jfrog_agent/slack_bot.py, line 10
4. Install to your workspace; copy xoxb- and xapp- tokens into .env
Agents/jfrog-agent/jfrog_agent/slack_bot.py, line 147
"Set SLACK_BOT_TOKEN (xoxb-…) and SLACK_APP_TOKEN (xapp-…) in .env"
Agents/jfrog-agent/jfrog_agent/context_optimizer/__init__.py, line 145
api_key = os.getenv("OPENAI_API_KEY")
Agents/jfrog-agent/jfrog_agent/context_optimizer/__init__.py, line 241
(kept so old .env files still work)
Agents/jfrog-agent/jfrog_agent/tracking/bq_client.py, line 56
key_path = os.environ.get("BIGQUERY_CREDENTIALS_PATH") or os.environ.get(
Agents/jfrog-agent/scripts/demo_context_optimizer_ab.py, line 75
max_tok = os.environ.get("CTX_OPT_MAX_TOKENS", "8000")
Agents/jfrog-agent/ui/state.py, line 34
Re-reads .env on each call so a long-running Streamlit process picks up
agent-safety-kit/agent_safety/input_guard/secrets.py, line 126
r"id_rsa|id_ed25519|id_ecdsa|"
agent-safety-kit/agent_safety/input_guard/secrets.py, line 128
r"\.env(?:\.\w+)?|"
agent-safety-kit/examples/01_guard_text.py, line 10
"here is my credentials.json contents",
artifactory/server.py, line 27
ARTIFACTORY_ACCESS_TOKEN = os.getenv("ARTIFACTORY_ACCESS_TOKEN")
Agent instructions
Pass
No instruction to ignore earlier instructions or to hide a step.
Obfuscation
Note
a file was over the read cap and was not fully read.
OBFUSCATION_FILE_OVER_READ_CAP
mcp-gateway/package-lock.json, line 1
{
Licence
Pass
A licence file is present (LICENSE).
Maintainer activity
Pass
The latest commit is within the last 365 days.
Note
Hygiene check of commit f9698059ef79. requests shell access. No manifest field names shell or command execution. install fetches a script from the network. can reach a host that is not written as a fixed address. reads a credential or environment file. a file was over the read cap and was not fully read.
Date and source
Date checked:
github.com
Commit graded: f9698059ef79a4e06681b16e7430e175d7f39792
This link goes to the original host. PlugVet does not host or check what you download from there, and installing it is your own decision and risk.