ClawHub

Auto Invoke Router

Security checks across malware telemetry and agentic risk

Overview

This skill transparently generates local skill-routing configuration and has no evidence of hidden execution, credential access, exfiltration, or destructive behavior beyond disclosed config replacement.

Before installing, be comfortable with the skill reading local skill descriptions and updating routing configuration. Review the generated triggers, especially broad or ambiguous ones, and use router.yml or back up AGENTS.md first if you want a manual approval step before changing future auto-invoke behavior.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The regeneration comment says "Regenerate by invoking: auto-invoke-router," but the document does not define whether this exact phrase is the activation trigger, nor does it provide exclusions or narrower invocation context. For a markdown skill description, this can create ambiguous routing or unintended activation because the phrase is a generic skill name rather than a clearly scoped command pattern.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill writes a generated routing section into AGENTS.md and later states that an existing section will be fully replaced, which can overwrite user-maintained configuration. Although overwrite details appear later in the instructions, the top-level description does not clearly warn users up front that running the skill modifies local files and may discard manual edits.

Self-Modification

High
Category
Rogue Agent
Content
triggers:
        - search clawhub
        - install skill
        - update skill
        - publish skill
        - clawhub list
    - skill: weather
Confidence
85% confidence
Finding
update skill

Session Persistence

Medium
Category
Rogue Agent
Content
Use the actual installed skill names and generated triggers — the above is illustrative only.

### Step 6 — Write output

**Option A — Append to AGENTS.md (recommended):**
Confidence
60% confidence
Finding
Write output **Option A — Append to AGENTS.md (recommended):** Add a `## Skill Routing` section at the end of `AGENTS.md` containing the full `skill_routing:` YAML block inside a fenced code block.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal