Most static findings are false positives caused by Markdown backticks around tool names, parameters, and a short hashing example. One low-severity network dependency is confirmed because the skill directs users to add the external Rube MCP endpoint. No prompt injection, hidden code execution, credential harvesting, or exfiltration intent was found in SKILL.md.
1
Files scanned
232
Lines analyzed
3
Review items
0
False positives ignored
Capability review items (1)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
**Get Rube MCP**: Add `https://rube.app/mcp` as an MCP server in your client configuration. No API k
Line 20 tells users to add https://rube.app/mcp as an external MCP server. This network dependency is intentional and disclosed, but Mailchimp workflow data may pass through a third party.
Most static findings are false positives caused by Markdown backticks around tool names, parameters, and a short hashing example. One low-severity network dependency is confirmed because the skill directs users to add the external Rube MCP endpoint. No prompt injection, hidden code execution, credential harvesting, or exfiltration intent was found in SKILL.md.
1
Files scanned
232
Lines analyzed
3
Review items
0
False positives ignored
Capability review items (1)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
**Get Rube MCP**: Add `https://rube.app/mcp` as an MCP server in your client configuration. No API k
Line 20 tells users to add https://rube.app/mcp as an external MCP server. This network dependency is intentional and disclosed, but Mailchimp workflow data may pass through a third party.
Most static findings are false positives caused by Markdown inline backticks around Mailchimp tool names and parameters. The only confirmed issue is a low-severity external MCP dependency at rube.app, which is disclosed by the setup instructions and appears necessary for the skill.
1
Files scanned
232
Lines analyzed
3
Review items
0
False positives ignored
Capability review items (1)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
**Get Rube MCP**: Add `https://rube.app/mcp` as an MCP server in your client configuration. No API k
Line 20 instructs users to add https://rube.app/mcp as an external MCP server. This is an expected network dependency, but it still routes Mailchimp workflows through a third-party service.
The static analyzer reported many shell execution findings, but review shows they are Markdown inline tool names and parameter names, not executable Ruby or shell code. The real risk is operational: the skill directs agents to use an external Rube MCP endpoint that can send campaigns, modify subscribers, and read Mailchimp analytics after OAuth authorization.
External MCP Service Can Perform Mailchimp Actions
TRUE_POSITIVE: The skill instructs users to add https://rube.app/mcp and authenticate a Mailchimp connection. This is expected functionality, but it sends requests through an external MCP service and can operate on a live Mailchimp account. Confidence: 0.86. Confidence reasoning: The endpoint and OAuth connection workflow are explicit, and the skill purpose depends on external Mailchimp access.
The external endpoint and OAuth connection workflow are explicit. The risk is legitimate integration exposure, not confirmed malicious behavior.
TRUE_POSITIVE: The workflow includes MAILCHIMP_SEND_CAMPAIGN and warns that sending is irreversible. A mistaken or unauthorized invocation could send email to an audience, so human approval is required before live sends. Confidence: 0.88. Confidence reasoning: The skill directly documents immediate sending and the irreversible nature of this action.
The live-send workflow and irreversible warning are explicitly documented. The concern is operational impact rather than malware.
TRUE_POSITIVE: The skill can add, update, search, and list subscribers, and it can retrieve campaign reports and activity. These actions may expose or modify personal data in Mailchimp, so users need correct authorization and scope control. Confidence: 0.84. Confidence reasoning: The documented tool sequences include subscriber upserts, searches, member lists, reports, and per-subscriber activity.
The data access and modification operations are clearly listed. They are normal Mailchimp tasks but involve sensitive contact data.
Needs review findings (1)
These findings came from uncertain legacy audit verdicts, so they require review but are not counted as confirmed security issues.
MD5 Subscriber Hash Is Not Protective Cryptography
NEEDS_REVIEW: The skill instructs users to compute an MD5 hash of a lowercase email for Mailchimp subscriber_hash values. This appears to follow Mailchimp API requirements, not to protect secrets, but MD5 email hashes should still be treated as personal data. Confidence: 0.72. Confidence reasoning: The MD5 usage is explicit and context shows it is an API identifier rather than a security control.
The weak hash use is explicit, but the surrounding text indicates a Mailchimp-required identifier. This is privacy-sensitive, not evidence of malicious cryptography.
Static false positives ignored (2)
These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.
Markdown Tool Names Misclassified as Shell Execution
FALSE_POSITIVE: The reported Ruby or shell backtick execution findings are inline Markdown references to Rube and Mailchimp tool names, parameters, or examples. No executable shell, Ruby backtick expression, or user-input command construction was found in SKILL.md. Confidence: 0.93. Confidence reasoning: The flagged locations are prose, lists, and tables inside a Markdown skill file.
The evidence is Markdown documentation with inline code formatting, not executable code. The file contains no script runner or shell invocation.
System Reconnaissance Findings Are Mailchimp Lookups
FALSE_POSITIVE: The flagged reconnaissance-like entries refer to Mailchimp list, segment, campaign, and member lookup workflows. They inspect Mailchimp account data after authorization and do not enumerate the local system. Confidence: 0.89. Confidence reasoning: The surrounding context is Mailchimp API workflow documentation, with no host reconnaissance commands.
The flagged lines describe Mailchimp object lookup or validation. No local host or network scanning behavior is present.
Static analysis detected 189 external_commands patterns and 1 network reference, but all are false positives. The skill is documentation for MCP (Model Context Protocol) tool calls to Mailchimp API, not actual shell commands. Network reference is the legitimate Rube MCP endpoint configuration. No malicious intent detected. Safe for publication.