Back to skill

Security audit

Tech Design Doc Generator

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill benign with medium confidence: the reported research workflow fits its technical documentation purpose, though it may send proposal keywords to search services or a configured Dify service, and one referenced path could not be resolved.

Before installing, consider whether proposal keywords might include confidential project details and whether your search and Dify services are approved for them. The scan is partial because a referenced local path was missing, so the benign verdict is not a guarantee that every reference was inspected.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

The research workflow sends proposal keywords to search services and a configured Dify API without a privacy warning.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The workflow directs the agent to submit proposal-derived keywords to web search and a configured retrieval service without distinguishing public information from confidential project details. In a technical architecture workflow, those keywords could expose internal product names, planned capabilities, or architecture choices to services outside the project's trust boundary. This is a conditional confidentiality risk, not evidence of deliberate exfiltration; an approved internal Dify deployment would reduce that portion of the risk.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 · SKILL.md (reported line 41)May include surrounding context.

如果候选方案信息不足:

  • 使用 web_search + web_fetch 检索候选方案资料
  • 用 exec 执行 curl 调用 Dify 检索 API:
    bash
    curl -s -X POST "${DIFY_API_BASE}/v1/datasets/${DIFY_DATASET_ID}/retrieve" \
      -H "Authorization: Bearer ${DIFY_API_KEY}" \