Back to skill

Security audit

Youtube Comment Manager

Security checks for vulnerabilities and agentic risk

Overview

The skill openly manages YouTube comments, but it combines real posting, deletion, moderation, and banning authority with mutable setup and live-instruction paths that users should review carefully.

Install only if you are comfortable giving AgentPMT-backed tooling authority over YouTube comments for the connected Google account. Before use, require explicit approval for posting replies, deleting comments, rejecting comments, bulk moderation, and any ban_author action, and prefer pinned or reviewed setup-skill versions rather than floating install commands.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:263
Finding

Unpinned Third-Party Skill Installation and Mutable Remote Instructions

Content
View full analysis

Vulnerability Details

File Location: SKILL.md:263-289, SKILL.md:306-324
Vulnerability Type: Supply-chain risk through unpinned remote dependencies and mutable operational instructions
Risk Level: Medium

Vulnerable Code Snippets

SKILL.md:263-289:

json
{
  "action": "get_instructions",
  "tool_id": "youtube-comment-manager"
}
json
{
  "name": "agentpmt-tool-search-and-execution",
  "parameters": {
    "action": "get_schema",
    "tool_id": "youtube-comment-manager"
  }
}
json
{
  "name": "agentpmt-tool-search-and-execution",
  "parameters": {
    "action": "get_instructions",
    "tool_id": "youtube-comment-manager"
  }
}

The preceding instructions state that remotely returned schema and instructions are more specific than the locally audited summary.

SKILL.md:306-324:

markdown
If those setup skills are not installed beside this product skill, use the downloads below.

Core AgentPMT setup skills:
- What AgentPMT is: ../what-is-agentpmt
  - ClawHub page: https://clawhub.ai/agentpmt/what-is-agentpmt
  - OpenClaw install: `openclaw skills install what-is-agentpmt`
  - skills.sh install: `npx skills add AgentPMT/agent-skills --skill what-is-agentpmt`
- AgentPMT account MCP/REST setup: ../agentpmt-account-mcp-rest-api-setup
  - ClawHub page: https://clawhub.ai/agentpmt/agentpmt-account-mcp-rest-api-setup
  - OpenClaw install: `openclaw skills install agentpmt-account-mcp-rest-api-setup`
  - skills.sh install: `npx skills add AgentPMT/agent-skills --skill agentpmt-account-mcp-rest-api-setup`

skills.sh install script:

```bash
npx skills add AgentPMT/agent-skills --skill what-is-agentpmt
npx skills add AgentPMT/agent-skills --skill agentpmt-account-mcp-rest-api-setup
text

### Technical Analysis

The Skill directs users to install third-party setup skills through `npx`, OpenClaw, or ClawHub without pinning a pa
...[truncated 2798 chars]
Remediation
View remediation

Remediation Suggestions

  1. Pin every package and setup skill to a reviewed version or immutable commit hash rather than resolving the latest upstream content.
  2. Publish SHA-256 checksums or signed manifests and verify them before installation.
  3. Avoid npx execution of mutable packages where possible. Vendor reviewed setup material or use a locked dependency installation process with integrity metadata.
  4. Separate remote schemas from trusted instructions. Treat all remotely returned text as untrusted data that cannot override local safety policy, user intent, authorization requirements, or confirmation gates.
  5. Permit remote schema updates only after validating them against a locally defined allowlist of action names, fields, destinations, and privilege boundaries.
  6. Display the resolved package version, source, checksum, and requested permissions before installation, and require explicit user approval.
  7. Run setup components with minimum privileges in a sandbox that has no unnecessary filesystem, credential, network, or tool access.
  8. Require explicit confirmation before destructive YouTube operations, especially bulk rejection, comment deletion, and ban_author: true.
  9. Maintain an auditable update process so that changes to dependencies and hosted instructions trigger a new security review.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (12)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill encourages moderation workflows including rejection and author banning, but the usage guidance does not consistently require an explicit user confirmation checkpoint before destructive or irreversible actions. Given that ban_author is noted elsewhere as irreversible from this tool, insufficient warning in workflow guidance raises the risk of accidental account moderation or permanent bans.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The activation guidance is broad enough that an agent may invoke this skill in loosely related contexts involving YouTube, comments, moderation, or channel management. Because the skill supports write and irreversible moderation actions, overbroad triggering can cause unintended external actions against real user accounts or comments.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

The skill instructs users to install supporting skills via npx skills add ... without pinning an exact package/repository version or immutable commit. That creates a supply-chain risk: if the upstream package or referenced skill changes, users may fetch unexpected or malicious content at install time.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

This line again recommends npx skills add ... without version pinning, which means the actual code or skill definition retrieved later may differ from what was reviewed. In an agent tooling ecosystem, that can silently introduce new behavior, permissions, or prompts.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

The install script uses npx skills add without a pinned version, exposing consumers to dependency drift and possible upstream compromise. Because this skill explicitly chains trust to other setup skills, unpinned retrieval expands the attack surface beyond this file.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

This repeated unpinned install command has the same supply-chain weakness: users may receive different code over time than what security reviewers examined. A compromised publisher, package, or dependency path could lead to execution of untrusted setup logic.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

The reference section again promotes npx skills add with no version constraints, which is a real supply-chain risk rather than just documentation noise. Repetition increases likelihood of users following the unsafe install path.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

Another unpinned npx skills add command allows the installed skill contents to change after review, undermining reproducibility and trust. In security-sensitive agent setups, mutable install targets are dangerous because they can alter remote tool behavior or setup flows.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 376)May include surrounding context.

md
- What AgentPMT is: ../what-is-agentpmt (ClawHub: `what-is-agentpmt`, page: https://clawhub.ai/agentpmt/what-is-agentpmt; skills.sh: `npx skills add AgentPMT/agent-skills --skill what-is-agentpmt`)
- AgentPMT account MCP/REST setup: ../agentpmt-account-mcp-rest-api-setup (ClawHub: `agentpmt-account-mcp-rest-api-setup`, page: https://clawhub.ai/agentpmt/agentpmt-account-mcp-rest-api-setup; skills.sh: `npx skills add AgentPMT/agent-skills --skill agentpmt-account-mcp-rest-api-setup`)
- Marketplace product: https://www.agentpmt.com/marketplace/youtube-comment-manager
- AgentPMT main MCP server: https://api.agentpmt.com/mcp/
- AgentPMT REST invoke endpoint: https://api.agentpmt.com/products/purchase

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 377)May include surrounding context.

md
- What AgentPMT is: ../what-is-agentpmt (ClawHub: `what-is-agentpmt`, page: https://clawhub.ai/agentpmt/what-is-agentpmt; skills.sh: `npx skills add AgentPMT/agent-skills --skill what-is-agentpmt`)
- AgentPMT account MCP/REST setup: ../agentpmt-account-mcp-rest-api-setup (ClawHub: `agentpmt-account-mcp-rest-api-setup`, page: https://clawhub.ai/agentpmt/agentpmt-account-mcp-rest-api-setup; skills.sh: `npx skills add AgentPMT/agent-skills --skill agentpmt-account-mcp-rest-api-setup`)
- Marketplace product: https://www.agentpmt.com/marketplace/youtube-comment-manager
- AgentPMT main MCP server: https://api.agentpmt.com/mcp/
- AgentPMT REST invoke endpoint: https://api.agentpmt.com/products/purchase

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

This markdown schema describes a delete_comment action that deletes a YouTube comment, but it provides no warning that the operation is destructive or potentially irreversible. For markdown files, safety-impacting behaviors affecting user data or system integrity should include an explicit warning or cautionary note.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The moderate_comments action allows changing moderation state and optionally banning an author, which can materially affect user accounts and content visibility. The schema lists parameters but does not warn users about the impact of rejecting comments or banning authors.

Content

No source excerpt is available for this finding.