Back to skill

Security audit

Hookbird autopilot

Security checks for vulnerabilities and agentic risk

Overview

ClawScan labels this skill benign: the scans found no high-risk behavior, while noting an agency-related warning and incomplete reference checks.

The overall verdict is benign, with medium confidence. Before installing, consider that two references to hookbird-log.md were not found in the scanned artifact, so checks involving those references are incomplete; the scans do not establish that the skill is risk-free.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 69)May include surrounding context.

md
- Only Studio spends credits; hooks, brand, plans, Radar, benchmarks, post stats and sales are free.
- One approval per week covers the posts and the credit total it showed. Anything new needs a new yes.
- Never publish without approval, never exceed the monthly budget, never make health, money or result claims the
  brand cannot prove.
- Judge a hook on 2-3 posts, never one: the same post can get 100x different views by chance.
- If a tool says a source is off, tell the owner why instead of guessing.