ClawHub
Back to skill

Security audit

Narrative Cascade Planner

Security checks for vulnerabilities and agentic risk

Overview

This skill is a scoped marketing planning helper that reads and writes local narrative memory with user confirmation and does not show hidden or destructive behavior.

Before installing, be aware that the skill can read your local narrative and claims memory and can save planning outputs if you confirm. It is intended for owned marketing surfaces; do not use it on private or third-party content unless you are comfortable with that content being analyzed for message consistency.

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
Content
## Save Results

After delivering the cascade plan, ask: "Save these results for future sessions?" On confirmation, write `memory/narrative/narrative-cascade-planner/YYYY-MM-DD-<topic>.md` per the [Skill Contract](../../../references/skill-contract.md) §Save Results Template. Referenced-but-unapproved claim wording goes only to `memory/events/claims.ndjson` via an authorized `operation: propose` request to `registry-events.py`; any canon-grade positioning/voice fact that surfaces goes only to `memory/events/narrative.ndjson` via an authorized `operation: propose` request to `registry-events.py` — narrative-registry owns the canon record. Do not write memory without asking.

## Reference Materials
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.

Static analysis

No suspicious patterns detected.