ClawHub
Back to skill

Security audit

JishuDB Knowledge Search / JishuDB知识库检索助手

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed, read-only helper for searching an authorized JishuDB knowledge base, with no evidence of hidden persistence, destructive behavior, or unrelated data transfer.

Install this only where the configured JISHUDB_URL is a trusted JishuDB backend and any JISHUDB_TOKEN is scoped appropriately. Expect the skill to read/search knowledge-base content and return citations, but it does not itself import data or modify the knowledge base.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The implemented behavior is narrower than the declared description in some areas and partially aligned in others. It does perform JishuDB-only access rather than public-web research, and it supports KB search plus listing available knowledge bases. However, the code only calls /api/kb for KB listing and /api/kb//search for semantic search results. There is no code for enumerating files within a KB, fetching full document bodies, or querying import/progress/status endpoints. So the description overstates the skill’s capabilities beyond what this code chunk actually does.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding

The skill instructs use of shell and environment-based HTTP access (bash, curl, JISHUDB_URL, JISHUDB_TOKEN) but does not declare an explicit tool/permission scope. That creates an authorization and review gap: a host may expose broader shell or env access than intended, enabling unintended data access or command execution paths beyond the documented read-only purpose.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · kb-search.sh (reported line 56)May include surrounding context.

sh
payload="{\"query\":\"${QUERY//\"/\\\"}\",\"topK\":${TOPK}}"
fi

resp=$(curl -sS --max-time 40 "${AUTH[@]}" -H 'content-type: application/json' \
  -X POST "$BASE/api/kb/$KB/search" -d "$payload") \
  || die "cannot reach KB at $BASE — is the 知识库 app running?"

Static analysis

No suspicious patterns detected.