ClawHub
Back to skill

Security audit

consensus-interact

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed consensus-workflow helper with local-first defaults and optional hosted use, with no evidence of hidden or destructive behavior.

Install only if you want an agent to manage consensus.tools jobs. Keep local mode and optional mutating tools disabled unless you deliberately want submissions, votes, resolutions, or hosted-board network actions. Treat hosted mode as a third-party data boundary and do not send secrets or sensitive job artifacts without approval.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Self-Modification

High
Category
Rogue Agent
Confidence
90% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · AI-SELF-IMPROVEMENT.md (reported line 34)May include surrounding context.

md
Consensus results are recorded and compared across iterations, allowing agents to track which decision patterns actually improved outcomes rather than merely sounding coherent.

3. **Governance for Self-Modification**  
   Instead of allowing unrestricted self-updates, consensus.tools introduces explicit validation before changes to goals, weights, or behavior are adopted.

4. **Mathematical Rather Than Narrative Rigor**  
   Decisions are selected through consensus mechanisms and scoring, not by whichever response is most fluent or confident.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill provides concrete instructions for switching to a hosted remote board, including setting an API key and submitting/voting on jobs, but it does not warn that job inputs, artifacts, summaries, and related metadata may be transmitted to an external service. In an agent-skill context, this can cause inadvertent disclosure of sensitive task data or credentials because the operator may follow the documented remote path without understanding the trust boundary change from local-first to hosted processing.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.