Back to plugin

Security audit

ClawFlow

Security checks for vulnerabilities and agentic risk

Overview

ClawFlow appears to be a legitimate workflow plugin, but it needs Review because flows can run shell commands, call agents and external APIs, persist sensitive state, and be configured for unattended full-exec operation.

Install only if you intend to let OpenClaw author and run automations. Keep the approval gate enabled unless the agent is isolated, inspect flows before running them, avoid unattended full-exec agents for untrusted inputs, do not expose the optional HTTP flow server without external access controls, and avoid putting long-lived secrets into flow state or persisted env values.

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill description uses broad phrases like creating a workflow, automation, pipeline, or flow, which overlap with many common user intents and can cause the skill to activate in situations beyond its safest intended scope. Over-broad activation increases the chance an agent will enter a high-capability workflow-authoring context that enables execution, HTTP calls, agent delegation, and environment access without sufficiently specific user intent.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The 'When to use this' section contains ambiguous invocation conditions such as 'user wants to chain AI calls' or 'user says create a flow,' without clear guardrails for when not to invoke the skill. In this context, ambiguity is risky because the skill exposes dangerous capabilities including shell execution, external HTTP requests, delegated agents with unrestricted exec configuration, and persistent memory operations.

Content

No source excerpt is available for this finding.

Endor dependency reachability

Checked Nov 9, 2026 · 10:22 PM UTC

No reachable vulnerabilities reported

Endor dependency scan completed.