ClawHub
Back to skill

Security audit

Industry Report Sprint / 行业研究报告速写

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed industry-report workflow that writes local report files and approved JishuDB records, with explicit user approval and credential-handling safeguards.

Install this if you are comfortable letting the agent use approved browsing/search, create local report files, and save research records into a selected JishuDB knowledge base. Confirm the KB, retention scope, output directory, and geography before use, especially if the report should not default to mainland China or should not retain certain source material.

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
  • Rogue AgentSelf-Modification, Session Persistence
  • 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 (2)

Self-Modification

High
Category
Rogue Agent
Content
1. Create a new `industry-report-sprint-<runId>` directory under the user's
   output directory or the host workspace location approved above. Do not
   overwrite existing files. Retain run ID, revision, stable record UUIDs,
   creation timestamps, and source/upload retry identities.
2. Retrieve relevant approved prior evidence with
   `kb_search({kbId, query, topK: 5})`, then
Confidence
85% 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.

Natural-Language Policy Violations

Low
Confidence
95% confidence
Finding
The instructions state defaults including "mainland China as the geographic starting point," which imposes a locale/geography assumption unless the user refines it. This is a natural-language policy concern because the skill applies a region-specific default rather than clearly obtaining user opt-in for the locale/context.

Static analysis

No suspicious patterns detected.