commercial-legal
Not malicious, needs a look Reviewed: false alarm
Bawwab's static rules found patterns that deserve a human look. This is not a finding that the package is malicious; the review notes below say what each pattern turned out to be.
Why it was flagged
The static rules marked this “Not malicious, needs a look” because of:
fetch.failed: Target could not be fetched (med, 2×)injection.html-comment: Hidden instruction in an HTML comment (med, 2× · first at skills/cold-start-interview/SKILL.md:125)
- Kind
- Plugin
- Source
- github
- Repository
- anthropics/claude-for-legal / commercial-legal
- Commit
4a6c651889c97cc9140580363c73e0eb17379c2b- Listed in
- anthropics/claude-for-legal
- Popularity
- 10,050 GitHub stars (2026-11-20)
- Link
- https://github.com/anthropics/claude-for-legal/tree/4a6c651889c97cc9140580363c73e0eb17379c2b/commercial-legal
- Scanned
- 2026-11-20T23:15:32+00:00
- Scan time
- 11.2 s
- Scanned as
- plugin
Review note: The HTML comments are harmless authoring notes. Two listed remote connectors could not be checked: TopCounsel's endpoint returned 404, and Definely's host (mcp.uk.definely.com) no longer resolves in DNS, so its tools were not scanned, and a lapsed domain could be registered by someone else.
Findings (4)
injection.html-comment: Hidden instruction in an HTML comment med
skills/cold-start-interview/SKILL.md:125
OWASP LLM LLM01 Prompt InjectionOWASP Agentic ASI01 Agent Goal HijackMITRE ATLAS AML.T0051.001 LLM Prompt Injection: IndirectMITRE ATLAS AML.T0068 LLM Prompt ObfuscationCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting
<!-- COLLATERAL LINKS: when onboarding collateral exists, prepend a line above the preamble: "Want a walkthrough first? [Watch the 3-minute intro](URL) or [read the getting-started guide](URL), then c
What this means: HTML comments are invisible when a document is rendered but are still read by the AI, so they can carry instructions meant to stay unseen.
The rule's general advice: Read the comment; remove it or do not install. (see the review note for this case)
Known false alarm
An authoring note for maintainers about adding onboarding links once they exist. It does not steer the agent toward anything harmful.
Reviewed by Claude (AI-assisted review), 2026-11-21
injection.html-comment: Hidden instruction in an HTML comment med
skills/stakeholder-summary/SKILL.md:87
OWASP LLM LLM01 Prompt InjectionOWASP Agentic ASI01 Agent Goal HijackMITRE ATLAS AML.T0051.001 LLM Prompt Injection: IndirectMITRE ATLAS AML.T0068 LLM Prompt ObfuscationCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting
<!-- Do not claim "I've added it to the tracker" unless `renewal-tracker` has actually been run for this contract — see Verify tracker entries before asserting them below. -->
What this means: HTML comments are invisible when a document is rendered but are still read by the AI, so they can carry instructions meant to stay unseen.
The rule's general advice: Read the comment; remove it or do not install. (see the review note for this case)
Known false alarm
A note telling the model not to claim a contract was added to the tracker unless the tracker skill actually ran. It is an honesty guard, not a hidden harmful instruction.
Reviewed by Claude (AI-assisted review), 2026-11-21
fetch.failed: Target could not be fetched med
[server TopCounsel] remote MCP server unreachable or invalid (http: ExceptionGroup: unhandled errors in a TaskGroup (1 sub-exception); sse: HTTPStatusError: Client error '404 Not Found' for url 'https://api.techgc.co/api/mcp/topcounsel' For more information check: https://developer.mozilla.org/en-U…
What this means: The package or server could not be downloaded, so nothing was reviewed.
The rule's general advice: Check the name, version and network, then retry. (see the review note for this case)
Known false alarm
Not a security finding: the TopCounsel connector (api.techgc.co) returned 404 when the scanner tried it, likely because it needs a login. Its tool list was not reviewed.
Reviewed by Claude (AI-assisted review), 2026-11-21
fetch.failed: Target could not be fetched med
[server Definely] remote MCP server unreachable or invalid (http: ExceptionGroup: unhandled errors in a TaskGroup (1 sub-exception); sse: ConnectError: [Errno -2] Name or service not known)
What this means: The package or server could not be downloaded, so nothing was reviewed.
The rule's general advice: Check the name, version and network, then retry. (see the review note for this case)
Known false alarm
Not a security finding: the Definely connector host (mcp.uk.definely.com) did not resolve in DNS when scanned, so its tools were not reviewed. Worth rechecking, since a configured domain that no longer exists could later be registered by someone else.
Reviewed by Claude (AI-assisted review), 2026-11-21
5 low-severity notes (informational, do not affect the result)
plugin.bundled-mcp: Plugin bundles an MCP server low
.mcp.json
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0010.005 Supply Chain Compromise: AI Agent Tool
plugin bundles MCP server(s): Definely, DocuSign, Google Drive, Ironclad, Slack, TopCounsel, iManage
What this means: The plugin installs additional MCP servers that run code on your machine and are not covered by this layer alone.
What you can do: Review each bundled server as if you were installing it separately.
fetch.auth-required: Remote server requires sign-in low
[server Ironclad] remote server answered HTTP 401; sign-in required
What this means: The remote server requires sign-in; its tool descriptions could not be read. Bawwab re-checks at session start.
What you can do: Sign in through Claude; Bawwab re-checks the tool descriptions at session start.
fetch.auth-required: Remote server requires sign-in low
[server DocuSign] remote server answered HTTP 401; sign-in required
What this means: The remote server requires sign-in; its tool descriptions could not be read. Bawwab re-checks at session start.
What you can do: Sign in through Claude; Bawwab re-checks the tool descriptions at session start.
fetch.auth-required: Remote server requires sign-in low
[server iManage] remote server answered HTTP 401; sign-in required
What this means: The remote server requires sign-in; its tool descriptions could not be read. Bawwab re-checks at session start.
What you can do: Sign in through Claude; Bawwab re-checks the tool descriptions at session start.
fetch.auth-required: Remote server requires sign-in low
[server Slack] remote server answered HTTP 401; sign-in required
What this means: The remote server requires sign-in; its tool descriptions could not be read. Bawwab re-checks at session start.
What you can do: Sign in through Claude; Bawwab re-checks the tool descriptions at session start.
Dispute this result
Maintainer and think this is wrong? Open a dispute on GitHub with the target id plugin-anthropics-claude-for-legal-commercial-legal-07f6a0.