Back to skill

Security audit

Tmp.EiWSQt4h8g

Security checks for vulnerabilities and agentic risk

Overview

The skill is read-only and aligned with homes.com lookups, but it asks users to run an unpinned npm MCP and install an external Chrome extension that works through a signed-in browser session.

Install only if you are comfortable trusting the homes-mcp npm package and the separate fetchproxy extension. Prefer pinning and reviewing exact versions, and use a dedicated Chrome profile with only the homes.com session needed for this tool. Treat saved homes and saved searches as private account data visible to the MCP flow.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T08 · Insecure Dependencies

Error
Location
SKILL.md:19
Finding

Unpinned npm Execution and Unverified Browser Extension Create Supply-Chain Risk

Content
View full analysis
Remediation
View remediation
"] } } } ``` 2. Prefer a locked local installation using a committed lockfile and npm integrity metadata. Review the package and transitive dependency changes before upgrading. 3. Avoid suppressing installation confirmation where practical. Separate dependency installation from normal Skill execution so upgrades cannot silently introduce new code. 4. Pin the `fetchproxy` extension to a reviewed release tag and immutable commit hash. Publish cryptographic checksums for release artifacts and document how users can verify them. 5. Provide reproducible extension build instructions and include the reviewed manifest and required permissions in the Skill documentation. 6. Minimize extension permissions to homes.com and the minimum local bridge endpoints needed for operation. Avoid broad host permissions, cookie access, or access to unrelated tabs unless strictly necessary. 7. Recommend a dedicated Chrome profile containing no unrelated authenticated sessions, extensions, or sensitive browsing data. 8. Make authenticated-session use optional for public listing operations. Request an authenticated session only when the user explicitly invokes saved-home or saved-search functionality. 9. Document the categories of browser and account data visible to the MCP server and extension, their network destinations, and whether any data is retained. 10. Add a dependency review and release-verification process covering npm ownership, provenance attestations, automated vulnerability scanning, and extension permission changes. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

MCP Config Access

High
Category
Agent Snooping
Confidence
95% confidence
Finding

Skill accesses MCP server configuration files (mcp.json). MCP configs contain server URLs, authentication tokens, and tool definitions — reading them allows the skill to discover and potentially abuse other tool integrations.

Content

Scanner excerpt · SKILL.md (reported line 19)May include surrounding context.

1. Install homes-mcp

.mcp.json (project) or ~/.claude/mcp.json (global):

json
{

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The trigger text is overly broad and can cause the skill to activate for many generic real-estate requests involving homes.com, prices, history, or photos without clearly signaling that browser-mediated network access may occur. Over-broad activation increases the chance of unintended invocation of a tool that operates through a signed-in browser session, which can surprise users and expose session-scoped data flows they did not explicitly consent to in that moment.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill describes operation through a signed-in homes.com tab and includes authenticated saved-homes and saved-search features, but it does not prominently warn users that invoking those features may access data from their existing browser session. This can lead to inadvertent disclosure of private account information or confusion about the trust boundary between the agent, the extension, and the authenticated website session.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.