ClawHub
Back to skill

Security audit

Campfire Prediction

Security checks across malware telemetry and agentic risk

Overview

This skill is transparent about being an autonomous Campfire prediction-market agent, but it needs Review because it can place bets with stored credentials and includes under-scoped safety controls.

Install only if you intentionally want an autonomous Campfire prediction-market agent. Require manual approval for prediction publishing and every order, set strict spending limits outside the skill, keep API keys and wallet files encrypted, verify downloaded files against trusted checksums, and do not use IP changes to work around registration limits.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The documentation states that registration returns an API key 'one-time only' but does not warn integrators to securely capture, store, and rotate/revoke it if exposed. In an autonomous trading agent context, loss or mishandling of that bearer token can lead to unauthorized access to profile, prediction, trading, and reward-claim endpoints.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The API reference documents endpoints that create orders and claim rewards without any explicit warning about financial risk, irreversible actions, authorization checks, or recommended safeguards such as confirmation, limits, and dry-run validation. Because this skill is for an autonomous prediction market agent, omission of such warnings increases the chance that integrators expose live trading and fund-affecting operations without adequate controls.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The trigger list includes broad natural-language phrases such as "browse markets," "make prediction," and "place bet," which are plausible in ordinary conversation and can cause unintended activation of this skill. Because this skill performs sensitive financial and account-related actions in a prediction-market/web3 context, accidental invocation could expose account state, initiate trading workflows, or steer an agent into handling API keys and wallet-signature flows when the user did not explicitly intend to use Campfire.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.