Bawwab: Top 302 scan

workspace-mcp

Flagged as possibly malicious Reviewed: expected behaviour

Bawwab's static rules matched a pattern it treats as critical. Read the review notes: a reviewed false alarm is marked as such.

Why it was flagged

The static rules marked this “Flagged as possibly malicious” because of:

Kind
MCP server
Source
pypi
Version
2.6.0
Popularity
60,257 PyPI downloads/week (2026-11-20)
Link
https://pypi.org/project/workspace-mcp/2.6.0/
Scanned
2026-11-20T22:02:41+00:00
Scan time
7.4 s
Review note: Google Workspace server: the hits are its own .env loading, a .env block-list, and honest tool descriptions; the 'malicious' flag does not hold up.

Findings (9)

credential-access: Touches credentials or secrets high

PKG-INFO:162

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

- **Sensitive path blocking** — local file reads default to the managed attachment directory, and `validate_file_path()` also blocks `.env` and `.env.*` files, the server's own credentials directory,

What this means: Reads files or environment holding keys, tokens or sessions.

The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)

Known false alarm

README text describing the server's protection that blocks reads of .env files and credential folders.

Reviewed by Claude (AI-assisted review), 2026-11-21

credential-access: Touches credentials or secrets high

fastmcp_server.py:13

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

dotenv_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), ".env")

What this means: Reads files or environment holding keys, tokens or sessions.

The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)

Expected for what this tool does

Loads the server's own .env file from its install folder at startup for its OAuth settings; this is how the server is configured.

Reviewed by Claude (AI-assisted review), 2026-11-21

credential-access: Touches credentials or secrets high

main.py:77

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

dotenv_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), ".env")

What this means: Reads files or environment holding keys, tokens or sessions.

The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)

Expected for what this tool does

Loads the server's own .env file next to main.py (Google OAuth client settings) at startup; that file configures the server and is used to talk to Google APIs.

Reviewed by Claude (AI-assisted review), 2026-11-21

exfil.combo: Reads credentials and makes network calls critical

main.py:77

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0037 Data from Local SystemMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0025 Exfiltration via Cyber Means

dotenv_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), ".env")

What this means: Credential access together with outbound requests in one file is the shape of data theft.

The rule's general advice: Do not install unless the flow is understood and benign. (see the review note for this case)

Expected for what this tool does

Loads the server's own .env file next to main.py (Google OAuth client settings) at startup; that file configures the server and is used to talk to Google APIs.

Reviewed by Claude (AI-assisted review), 2026-11-21

credential-access: Touches credentials or secrets high

core/utils.py:335

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

if any(part == ".env" or part.startswith(".env.") for part in path_parts):

What this means: Reads files or environment holding keys, tokens or sessions.

The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)

Known false alarm

Part of validate_file_path(): it refuses to read, upload or attach any .env file. Protective code.

Reviewed by Claude (AI-assisted review), 2026-11-21

credential-access: Touches credentials or secrets high

scripts/release.py:77

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

DOTENV_PATH = REPO_ROOT / ".env"

What this means: Reads files or environment holding keys, tokens or sessions.

The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)

Expected for what this tool does

Maintainer release script that reads the repo's own .env for a PyPI token to publish releases. Not used when the server runs.

Reviewed by Claude (AI-assisted review), 2026-11-21

credential-access: Touches credentials or secrets high

workspace_mcp.egg-info/PKG-INFO:162

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

- **Sensitive path blocking** — local file reads default to the managed attachment directory, and `validate_file_path()` also blocks `.env` and `.env.*` files, the server's own credentials directory,

What this means: Reads files or environment holding keys, tokens or sessions.

The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)

Known false alarm

Copy of the same README text in the package metadata.

Reviewed by Claude (AI-assisted review), 2026-11-21

tool-shadowing: Tool description changes how another tool is used high

tool modify_doc_text:6

OWASP LLM LLM01 Prompt InjectionOWASP Agentic ASI02 Tool Misuse and ExploitationOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0110.000 AI Agent Tool Poisoning: Definition and InstructionsMITRE ATLAS AML.T0011.002 User Execution: Poisoned AI Agent ToolCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting

instead of start_index

What this means: A description that names a different tool and tells the AI how to use it (or to use it instead) can quietly hijack that tool, for example by adding a recipient to emails.

The rule's general advice: Do not install; a tool should only describe itself. (see the review note for this case)

Known false alarm

The modify_doc_text description explains its own parameters (pass `heading` instead of `start_index`). It does not mention or redirect any other tool.

Reviewed by Claude (AI-assisted review), 2026-11-21

injection.override: Tries to override the AI's instructions high

tool send_gmail_message:12

OWASP LLM LLM01 Prompt InjectionOWASP LLM LLM07 System Prompt LeakageOWASP Agentic ASI01 Agent Goal HijackMITRE ATLAS AML.T0051.001 LLM Prompt Injection: IndirectMITRE ATLAS AML.T0110.000 AI Agent Tool Poisoning: Definition and InstructionsMITRE ATLAS AML.T0069.002 Discover LLM System Information: System PromptCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting

Never tell a user

What this means: Text read by the AI that says to ignore earlier instructions, reveal the system prompt or hide things from you is a classic prompt-injection attempt.

The rule's general advice: Do not install; the text is trying to steer the assistant against you. (see the review note for this case)

Known false alarm

The send_gmail_message description says the tool sends immediately and cannot schedule, so the AI must "never tell a user a message was scheduled". This keeps the AI honest with the user, the opposite of hiding things.

Reviewed by Claude (AI-assisted review), 2026-11-21

3 low-severity notes (informational, do not affect the result)

install-dir-access: Reads or writes Claude's install folders low

PKG-INFO:314

MITRE ATLAS AML.T0081 Modify AI Agent Configuration

ln -s "$(pwd)/skills/managing-google-workspace" ~/.claude/skills/managing-google-workspace

What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.

What you can do: Check which files it adds there and that you want them.

credential-access: Touches credentials or secrets low

tests/test_main_permissions_tier.py:61

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

env = os.environ.copy()

What this means: Reads files or environment holding keys, tokens or sessions.

What you can do: Confirm the tool needs this; otherwise do not install.

install-dir-access: Reads or writes Claude's install folders low

workspace_mcp.egg-info/PKG-INFO:314

MITRE ATLAS AML.T0081 Modify AI Agent Configuration

ln -s "$(pwd)/skills/managing-google-workspace" ~/.claude/skills/managing-google-workspace

What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.

What you can do: Check which files it adds there and that you want them.

Dispute this result

Maintainer and think this is wrong? Open a dispute on GitHub with the target id mcp-pypi-workspace-mcp-0838f2.