MCP, MCP server
Mcp Server Security Standard
The Model Context Protocol enables AI models to interact with external systems through tools, resources, and prompts.
Compatibility: Not stated.
Grade C
One Medium finding. Read the report before installing.
Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.
PERMISSION_SHELL_UNDECLARED
Why this grade
- PERMISSION_SHELL_UNDECLARED Medium
Runs shell commands that the manifest or description does not mention.
v0.1/controls/MCP-AUTHZ-03-least-privilege.md, line 148
subprocess.run(path) # Can execute anything
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.
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
v0.1/controls/MCP-AUTHZ-03-least-privilege.md, line 148
subprocess.run(path) # Can execute anything
v0.1/controls/MCP-EXEC-01-no-shell-execution.md, line 13
MCP servers MUST NOT use shell invocation (e.g., `os.system()`, `subprocess.run(shell=True)`, `child_process.exec()`) with user-controlled input. Use parameterized execution instea
v0.1/controls/MCP-EXEC-01-no-shell-execution.md, line 25
1. Code review showing use of `subprocess.run([cmd, arg1, arg2], shell=False)` or equivalent
Show more
v0.1/controls/MCP-EXEC-02-command-allowlisting.md, line 51
return subprocess.run([cmd_name] + args, shell=False)
v0.1/controls/MCP-EXEC-03-argument-separator.md, line 45
subprocess.run(["git", "log", "--", user_provided_path], shell=False)
v0.1/controls/MCP-EXEC-03-argument-separator.md, line 46
subprocess.run(["grep", "-r", "pattern", "--", user_provided_dir], shell=False)
v0.1/controls/MCP-EXEC-03-argument-separator.md, line 53
subprocess.run(["git", "log", user_provided_path], shell=False)
v0.1/controls/MCP-EXEC-03-argument-separator.md, line 56
subprocess.run(["rm", user_provided_file], shell=False)
Install and update
Pass
No install-time script found, and dependencies are pinned or absent.
Network
Concern
can reach mcp-server. The description does not name that host.
NETWORK_UNNAMED_HOST
v0.1/controls/MCP-DEPLOY-03-resource-limits.md, line 193
response = requests.post('http://mcp-server/tools/execute')
Secrets
Concern
reads a credential or environment file.
SECRETS_PATH_DOTENV
v0.1/controls/MCP-AUTHZ-04-rbac.md, line 382
resources: ['**/sensitive/**', '**/.env', '**/secrets.json'],
v0.1/research/notes.md, line 48
| [Kaspersky GERT - Supply Chain Attack](https://securelist.com/model-context-protocol-for-ai-integration-abused-in-supply-chain-attacks/) | PoC malicious MCP server stealing crede
Agent instructions
Pass
No instruction to ignore earlier instructions or to hide a step.
Obfuscation
Pass
No encoded or minified payload found.
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 c848011308f9. requests shell access. No manifest field names shell or command execution. can reach mcp-server. The description does not name that host. reads a credential or environment file.
Date and source
Date checked:
github.com
Commit graded: c848011308f94a38aec9a792179d7cda2d7ea9db
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.