Back to skill

Security audit

SideQuest: Hire Real People

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious because it reports an undisclosed $50 account top-up and gaps in how payment approval and sharing task details are explained.

Consider the reported $50 top-up and the 72-hour automatic payment release before installing. Check what task details and personal information would be sent to SideQuest and shared with a tasker, and whether the skill clearly explains those actions. The scan was partial because a local reference could not be resolved, so its evidence is incomplete.

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

Although account funding is related to a paid-task service, initiating a $50 top-up is a material financial action rather than a routine implementation detail. The description discloses hiring and escrow payments, but not adding funds to an account. The supplied code performs only the top-up request; it does not hire anyone or manage task escrow.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 11)May include surrounding context.

md
env:
        - SIDEQUEST_API_KEY
      bins:
        - curl
    primaryEnv: SIDEQUEST_API_KEY
    emoji: "\U0001F9ED"
    homepage: https://sidequestpays.com

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 60)May include surrounding context.

| be_present | Be somewhere for a set time (line, sign, table) with timed check-ins | Timestamped check-ins |

bash
curl https://sidequestpays.com/v1/templates

Before posting

Quest submission lacks a privacy warning about sharing detailed briefs with SideQuest and human questers.

Medium
Category
Output Handling
Confidence
88% confidence
Finding

The skill instructs the agent to collect exact locations, access details, and contact information and then submit a quest to an external service for a human to perform. Although it requests approval of a summary and cost, it does not warn that the detailed brief will be shared externally or advise against including sensitive information.

Content

No source excerpt is available for this finding.

Automatic approval after 72 hours contradicts the manifest's approval-gated escrow release.

Medium
Category
Excessive Agency
Confidence
98% confidence
Finding

The description promises escrow release when proof is approved, but the documented workflow also releases payment automatically after 72 hours. Although the timeout is disclosed later, the required user-facing confirmation omits it, potentially causing users to authorize spending under an incorrect understanding of their control over payment release.

Content

No source excerpt is available for this finding.