ClawHub
Back to skill

Security audit

Dark Social Attributor

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed marketing attribution workflow that reads user-provided analytics and only saves results or memory updates through stated, user-controlled paths.

Before installing, be aware that the skill is designed to process GA4/GSC exports, form inventories, and marketing-memory records. Use it with data you are comfortable analyzing in the agent environment, and only approve the optional save/propose steps when the generated facts are accurate.

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
  • 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
## Save Results

After delivering the pack, ask: "Save these results for future sessions?" On confirmation, save to `memory/social/dark-social-attributor/YYYY-MM-DD-<topic>.md` — see [Skill Contract](../../../references/skill-contract.md) §Save Results Template. Channel-grade facts go only to `memory/events/channels.ndjson` via an authorized `operation: propose` request to `registry-events.py` (channel-registry is the sole writer of `memory/channels/`); opt-in evidence goes to `memory/events/consent.ndjson` via an authorized `operation: propose` request to `registry-events.py`. Do not write memory without asking.

## Reference Materials