Pipeworx Housing Analyst

• Outcome: Review
• Latest audit: Sep 16, 2026 · 6:56 AM UTC
• Version: 1.0.0

Overview

The skill appears to do the advertised housing analysis, but it routes sensitive property queries through a third-party MCP gateway and encourages memory use without privacy limits.

Review this skill before installing if you plan to submit street addresses, ownership-linked property records, valuations, or coordinates. Treat the Pipeworx gateway as an external processor, avoid sending unnecessary personal or property-identifying details, and do not save property-specific findings in memory unless the user explicitly wants that retention.

A.I.G

Vulnerability Patterns

• Insecure Dependencies: Introduces malicious components through unsafe dependency sources
• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally

Findings (1)

T08 · Insecure Dependencies

Warning

• Location: SKILL.md:5

Finding

Unpinned Third-Party MCP Gateway Handles Housing and Property Queries

Content

View full analysis

Vulnerability Details

File Location: SKILL.md, lines 5-12; related data-processing capabilities are documented at lines 48-60
Vulnerability Type: T08: Insecure Dependencies
Risk Level: Medium

Vulnerable Code:

json

{
  "mcpServers": {
    "housing-intel": {
      "url": "https://gateway.pipeworx.io/mcp?task=housing%20analysis"
    }
  }
}

The configured gateway is used for capabilities involving property addresses and associated real-estate records:

markdown

| `attom_property_detail` | Get full property characteristics by address — lot size, square footage, bedrooms, bathrooms, year b |
| `attom_property_search` | Search properties by location with optional filters. Search by postal code or by latitude/longitude  |
| `attom_sales_history` | Get complete sales history for a property (up to 10 years) — sale dates, prices, deed types, seller/ |
| `attom_avm` | Get automated valuation (AVM) for a property — estimated market value, confidence score, value range |
| `attom_assessment` | Get property tax assessment details — assessed value, market value, tax amount, tax year, and assess |

Technical Analysis

The skill delegates all advertised FRED, BLS, ATTOM, and HUD functionality to the externally hosted gateway.pipeworx.io MCP endpoint rather than directly configuring the named official data providers. MCP servers can supply tool definitions and process arguments submitted through those tools. Consequently, this third-party service occupies a trusted supply-chain position and can receive property addresses, geographic coordinates, and housing-analysis parameters.

The configuration does not specify a pinned service or protocol version, integrity verification, an allowlist of expected tools, response-schema validation, data-retention controls, or privacy requirements. If the gateway or its upstream dependencies are compromised or changed, its behavior can differ from what wa ...[truncated 1757 chars]

Remediation

View remediation

Remediation Suggestions

1. Connect directly to official FRED, BLS, ATTOM, and HUD endpoints where practical, reducing reliance on an intermediary gateway.
2. Document the gateway operator, subprocessors, authentication model, retention period, logging practices, and privacy policy.
3. Obtain explicit user consent before transmitting street addresses, coordinates, or other property-identifying information.
4. Minimize submitted data and redact user-identifying details that are unnecessary for the requested analysis.
5. Pin a reviewed MCP protocol and service version or use an immutable, integrity-verified deployment where supported.
6. Allowlist the exact expected tool names and reject unexpected tools or material schema changes supplied by the remote server.
7. Validate response schemas, data types, provenance, and plausible value ranges before using results.
8. For consequential decisions, cross-check returned data against the named authoritative providers.
9. Establish contractual and technical retention limits and provide deletion controls for stored property-query data.
10. Monitor gateway certificate, ownership, tool-schema, and behavior changes, and disable access when unexpected changes are detected.

SkillSpector

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 94% confidence

Finding

The setup directs users to connect to a third-party MCP gateway URL but does not disclose that prompts, addresses, and analysis inputs will be transmitted to an external service. This creates a data-sharing and trust-boundary risk, especially because the listed tools handle potentially sensitive property and market queries that users may assume are local or first-party.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 89% confidence

Finding

The skill explicitly encourages use of property-analysis tools and remember/recall without warning that addresses, ownership history, valuations, and related property records can be sensitive when tied to individuals. Persisting such data in agent memory increases the chance of unnecessary retention, cross-task leakage, or disclosure to external systems beyond the user's expectation.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.