Back to skill

Security audit

yaps-video-clipping

Security checks for vulnerabilities and agentic risk

Overview

This appears to be a real Yaps video helper, but it exposes broader local Yaps actions than the video-clipping description clearly explains.

Review before installing. Use it only if you are comfortable with an agent invoking the local Yaps CLI, checking sanitized Yaps account status, probing the local Yaps installation, and possibly opening Yaps during account recovery. Keep tasks limited to the videos you intend to process and avoid giving it unrelated sensitive media or documents.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (20)

Context-Inappropriate Capability

High
Confidence
93% confidence
Finding
A general subprocess-backed execution path lets the skill act as a generic front-end to Yaps rather than a constrained video-clipping tool. In this skill context, that mismatch is more dangerous because the manifest promises a narrow editing action, but the runtime can invoke a much wider set of Yaps capabilities on user-supplied files and requests.

Description-Behavior Mismatch

High
Confidence
98% confidence
Finding
The adapter accepts arbitrary Yaps CLI arguments and forwards them to the installed Yaps binary, creating a broad command surface beyond the advertised video-clipping workflow. Even with shell injection mitigations, this is dangerous because it enables undisclosed operations against local files, account state, and Yaps features that the host and user did not explicitly authorize through the skill contract.

Credential Access

High
Category
Privilege Escalation
Content
function accountStatusUnsafeDiagnosis(appVersion) {
  return {
    code: "account_status_unsafe",
    message: `Yaps ${appVersion || "before 2.3.124"} uses an older credential-based account check, so this plugin deliberately did not run it. Update Yaps to 2.3.124 or newer; the plugin will then reuse the desktop sign-in, trial, or Yaps Pro automatically. Do not approve a Keychain prompt or create a separate plugin account.`,
  };
}
Confidence
70% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Credential Access

High
Category
Privilege Escalation
Content
function accountStatusUnsafeDiagnosis(appVersion) {
  return {
    code: "account_status_unsafe",
    message: `Yaps ${appVersion || "before 2.3.124"} uses an older credential-based account check, so this plugin deliberately did not run it. Update Yaps to 2.3.124 or newer; the plugin will then reuse the desktop sign-in, trial, or Yaps Pro automatically. Do not approve a Keychain prompt or create a separate plugin account.`,
  };
}
Confidence
70% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Credential Access

High
Category
Privilege Escalation
Content
function accountStatusUnsafeDiagnosis(appVersion) {
  return {
    code: "account_status_unsafe",
    message: `Yaps ${appVersion || "before 2.3.124"} uses an older credential-based account check, so this plugin deliberately did not run it. Update Yaps to 2.3.124 or newer; the plugin will then reuse the desktop sign-in, trial, or Yaps Pro automatically. Do not approve a Keychain prompt or create a separate plugin account.`,
  };
}
Confidence
70% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Credential Access

High
Category
Privilege Escalation
Content
function accountStatusUnsafeDiagnosis(appVersion) {
  return {
    code: "account_status_unsafe",
    message: `Yaps ${appVersion || "before 2.3.124"} uses an older credential-based account check, so this plugin deliberately did not run it. Update Yaps to 2.3.124 or newer; the plugin will then reuse the desktop sign-in, trial, or Yaps Pro automatically. Do not approve a Keychain prompt or create a separate plugin account.`,
  };
}
Confidence
70% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Credential Access

High
Category
Privilege Escalation
Content
message: "Yaps is signed in, but its trial or Yaps Pro access is not active. Open Yaps to review the available trial or Yaps Pro options; the plugin will pick up the change automatically.",
    };
  }
  if (auth.status === "credential_unavailable" || auth.diagnosticCode === "keychain_unavailable") {
    return {
      code: "account_status_unsupported",
      message: "This installed Yaps helper uses an older credential-based account check. Update Yaps and retry. Do not approve a Keychain prompt or create a separate plugin account.",
Confidence
70% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Lp3

Medium
Category
MCP Least Privilege
Confidence
84% confidence
Finding
The skill instructs use of local command-line tooling (`yaps`) and explicitly references environment/runtime behavior, but the manifest does not declare any `permissions` or `allowed-tools` scope. That mismatch weakens least-privilege controls and can cause the runtime to grant broader-than-expected execution or environment access, especially when the skill handles local files and rendering operations.

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The skill exposes transcription, subtitle generation, and meeting transcription workflows that materially exceed the declared purpose of shortening talking-head videos. Scope expansion is dangerous because users and hosts may grant trust or permissions based on the narrow manifest description, while the code can process and persist broader audio/video-derived content than expected.

Session Persistence

Medium
Category
Rogue Agent
Content
export const MIN_SAFE_AUTH_STATUS_VERSION = "2.3.124";
const MAX_PROBE_BYTES = 64 * 1024;
const MAX_VERSION_BYTES = 4 * 1024;
const MAX_PLIST_BYTES = 64 * 1024;
const MAX_WINDOWS_SHIM_BYTES = 4 * 1024;
const RUNNING_YAPS_PROCESS_NAMES = new Set(["yaps.exe", "yaps_mcp.exe", "yaps_cli.exe"]);
// Fixed WMI filter: never interpolate a discovered path into this command.
Confidence
75% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Session Persistence

Medium
Category
Rogue Agent
Content
export const MIN_SAFE_AUTH_STATUS_VERSION = "2.3.124";
const MAX_PROBE_BYTES = 64 * 1024;
const MAX_VERSION_BYTES = 4 * 1024;
const MAX_PLIST_BYTES = 64 * 1024;
const MAX_WINDOWS_SHIM_BYTES = 4 * 1024;
const RUNNING_YAPS_PROCESS_NAMES = new Set(["yaps.exe", "yaps_mcp.exe", "yaps_cli.exe"]);
// Fixed WMI filter: never interpolate a discovered path into this command.
Confidence
75% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Session Persistence

Medium
Category
Rogue Agent
Content
if (platform === "darwin") {
    const application = macApplicationForCli(cli.path, canonicalize);
    if (!application) return null;
    const plist = posix.join(application, "Contents", "Info.plist");
    try {
      const metadata = statFile(plist);
      if (!metadata.isFile() || metadata.size > MAX_PLIST_BYTES) return null;
Confidence
75% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Session Persistence

Medium
Category
Rogue Agent
Content
if (platform === "darwin") {
    const application = macApplicationForCli(cli.path, canonicalize);
    if (!application) return null;
    const plist = posix.join(application, "Contents", "Info.plist");
    try {
      const metadata = statFile(plist);
      if (!metadata.isFile() || metadata.size > MAX_PLIST_BYTES) return null;
Confidence
75% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Session Persistence

Medium
Category
Rogue Agent
Content
if (platform === "darwin") {
    const application = macApplicationForCli(cli.path, canonicalize);
    if (!application) return null;
    const plist = posix.join(application, "Contents", "Info.plist");
    try {
      const metadata = statFile(plist);
      if (!metadata.isFile() || metadata.size > MAX_PLIST_BYTES) return null;
Confidence
75% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Session Persistence

Medium
Category
Rogue Agent
Content
if (platform === "darwin") {
    const application = macApplicationForCli(cli.path, canonicalize);
    if (!application) return null;
    const plist = posix.join(application, "Contents", "Info.plist");
    try {
      const metadata = statFile(plist);
      if (!metadata.isFile() || metadata.size > MAX_PLIST_BYTES) return null;
Confidence
75% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Session Persistence

Medium
Category
Rogue Agent
Content
if (platform === "darwin") {
    const application = macApplicationForCli(cli.path, canonicalize);
    if (!application) return null;
    const plist = posix.join(application, "Contents", "Info.plist");
    try {
      const metadata = statFile(plist);
      if (!metadata.isFile() || metadata.size > MAX_PLIST_BYTES) return null;
Confidence
75% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Session Persistence

Medium
Category
Rogue Agent
Content
if (platform === "darwin") {
    const application = macApplicationForCli(cli.path, canonicalize);
    if (!application) return null;
    const plist = posix.join(application, "Contents", "Info.plist");
    try {
      const metadata = statFile(plist);
      if (!metadata.isFile() || metadata.size > MAX_PLIST_BYTES) return null;
Confidence
75% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Description-Behavior Mismatch

Medium
Confidence
91% confidence
Finding
The code performs session discovery and account-state inspection, including optional account recovery behavior, which is broader than the declared video-clipping workflow. Even if intended to support the product, this expands the agent's capability to inspect local app state and account information without clear disclosure, increasing privacy and trust risk.

Context-Inappropriate Capability

Medium
Confidence
95% confidence
Finding
The code can launch the installed Yaps desktop application in the background to refresh account state, which exceeds the obvious expectations of a video-clipping skill. Background app launching can trigger side effects, change user state, or be abused as an unintended execution primitive if the launched binary path validation is ever bypassed elsewhere.

Description-Behavior Mismatch

Low
Confidence
90% confidence
Finding
The skill enumerates running processes and probes local executables to locate Yaps components, but this host-inspection behavior is not reflected in the manifest description. While not directly code-execution, undisclosed process and filesystem discovery increases the attack surface and can surprise users or reviewers about the degree of local introspection being performed.

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical
Code
suspicious.dangerous_exec
Location
runtime/run.mjs:74