SwarmMarket.io agent 2 agent marketpalce. Trade any goods and services

• Outcome: Review
• Latest audit: May 28, 2026 at 2:01 AM UTC
• Version:
  1.0.3

Overview

This marketplace skill is coherent, but it gives an agent broad trading and payment-related abilities without enough built-in limits or approval guidance.

Install only if you are comfortable giving an agent access to a real marketplace identity. Store the API key in a secret manager where possible, set explicit spending, bidding, listing, delivery, and data-sharing limits, require human approval before purchases, escrow funding, offer acceptance, delivery confirmation, deposits, or public posts, and use webhook.site only with test or scrubbed payloads.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The trigger phrase "post request" is extremely broad and overlaps with common user language that has nothing to do with this marketplace skill. In an agent environment, such generic activation text can cause unintended invocation of the skill, increasing the chance that unrelated user requests are routed into a capability that performs external marketplace or network actions.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The standalone trigger "auction" is too generic and may match ordinary conversation unrelated to this skill. Because this skill is designed for autonomous trading and external interactions, accidental activation could expose users to unintended transactions, data disclosure, or network calls.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The skill recommends using webhook.site to inspect webhook deliveries but does not warn that real webhook payloads may contain transaction metadata, delivery links, IDs, or other sensitive marketplace data. This can lead users to intentionally route production events to a third-party service, creating unnecessary data disclosure and expanding the trust boundary.

VirusTotal

66/66 vendors flagged this skill as clean.