Back to skill

Security audit

ToolFlow OpenClaw Operator

Security checks for vulnerabilities and agentic risk

Overview

The skill has a coherent workflow purpose, but its bundled plugin can read arbitrary local files as a “safe” workflow action and persist those outputs, so it needs review before installation.

Install only if you are comfortable giving the plugin broad local-file read capability through workflows and storing workflow outputs in a local ledger. Keep elevated mode disabled unless needed, avoid command progress sinks, review workflow files before submitting them, and restrict the ledger/plugin configuration to a workspace you trust.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T05 · Unauthorized Access and Privilege Escalation

Error
Location
toolflow-bundle/plugin/dist/entry.js:1555
Finding

Unrestricted Local File Access in the Default Safe Profile

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
toolflow-bundle/plugin/dist/entry.js:929
Finding

Ledger Path Traversal Through Unvalidated Run Identifiers

Content
View full analysis
/runs` directory. The implementation does not: - Enforce the runtime's generated run-ID format. - Reject `/`, `\`, `.` or `..` path components. - Canonicalize the final path and verify that it remains under the ...[truncated 2453 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (19)

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/entry.js (reported line 412)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/entry.js (reported line 413)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/entry.js (reported line 414)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/entry.js (reported line 2121)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/entry.js (reported line 2198)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/entry.js (reported line 2199)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/runtime-cli.js (reported line 575)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/runtime-cli.js (reported line 576)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/runtime-cli.js (reported line 577)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/runtime-cli.js (reported line 2121)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/runtime-cli.js (reported line 2198)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · toolflow-bundle/plugin/dist/runtime-cli.js (reported line 2199)May include surrounding context.

js
}
});

// packages/runtime/dist/crypto/keyring.js
var require_keyring = __commonJS({
  "packages/runtime/dist/crypto/keyring.js"(exports2) {
    "use strict";

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The default progress reporter can spawn an arbitrary shell command from configuration/environment using shell: true, passing runtime data through environment variables. In this plugin context, that creates a hidden command-execution channel unrelated to the advertised tool action, so submitting or resuming workflows may trigger unintended code execution if configuration is influenced by an attacker or unsafe deployment.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The README instructs users to execute local shell scripts that install and verify a bundled plugin/runtime, modify the OpenClaw plugin setup, and run checks, but it provides no caution about reviewing the scripts, understanding side effects, or limiting privileges. This creates a social-engineering and supply-chain risk: users may run opaque installer code directly from the skill directory, which could alter local configuration or execute additional commands beyond what the README briefly describes.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The elevated apply_patch path writes attacker-controlled patch content to a temporary file and then applies it to an arbitrary resolved cwd via git apply, which can modify local files. In a skill/plugin context, this is dangerous because it enables filesystem mutation through a tool path without any user-facing warning or hard scoping to a safe workspace, increasing the risk of unauthorized code or config changes.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
77% confidence
Finding

The tool is described as only compiling and classifying a workflow without running it, but dryRun calls resolveConfig, which can source progress command settings from environment variables, and the runtime includes a default reporter that spawns shell commands when configured. This is an intent-level contradiction because the documentation promises non-execution semantics while the codebase for the invoked path still admits command execution side effects through its runtime configuration surface.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The elevated runner writes a temporary patch file and applies it with git apply, which can modify workspace files, but the function contains no confirmation prompt, warning comment, or visible logging to disclose that a write/destructive operation is occurring. Although approval is enforced elsewhere, this code path itself lacks direct user disclosure for the filesystem changes it performs.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

When progress reporting is configured with sink='command', the runtime spawns an arbitrary shell command from environment/config without any in-code consent gate at the operation point. Because this happens asynchronously during normal workflow execution, it creates an unexpected command-execution path that can be abused by anyone controlling environment variables or configuration.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The CLI presents the runtime as a 'Safe Profile' even though the same codebase can enable an elevated lane that executes commands and applies patches. This is a security-signaling flaw: operators may trust the tool as non-mutating or low-risk and enable or run it under false assumptions, increasing the chance of unsafe deployment or approval decisions.

Content

No source excerpt is available for this finding.