Back to skill

Security audit

Social Media Scheduler — Publish, Cross-Post & Auto-Post via ReplyNodes for OpenClaw

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed, confirmation-gated ReplyNodes social-posting helper; use pinned installation and confirm previews carefully because it can publish publicly after approval.

Install from a reviewed immutable tag or commit rather than the unpinned npx command. Before use, verify the connected ReplyNodes workspace and channels, review every preview, and only confirm a named run you actually want published or scheduled. Revoke the ReplyNodes session when uninstalling or if access is no longer needed.

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
README.md:12
Finding

Unpinned npx Installer and Mutable Repository Reference Enable Supply-Chain Execution

Content
View full analysis

Vulnerability Details

File Location: README.md:12 and README.md:135
Vulnerability Type: Unpinned third-party installer and mutable dependency source
Risk Level: Medium

Vulnerable Code

bash
npx skills add replynodes/replynodes-agent-skills --skill replynodes --agent openclaw --yes

The same command appears in both the installation and update instructions.

Technical Analysis

The documented command invokes the third-party skills package through npx without specifying an exact package version. Depending on the local npm cache and npx behavior, this can download and execute the version currently resolved by the npm registry. Consequently, the executable installer may differ from the version reviewed when this skill was audited.

The repository argument, replynodes/replynodes-agent-skills, is also not tied to an immutable release tag or full commit SHA in this command. Although the README separately provides a tagged direct-Git installation example, the primary npx installation and update commands do not provide equivalent source immutability.

This is a supply-chain risk rather than evidence that the current repository contains malicious code. Exploitation requires compromise or malicious alteration of the externally resolved installer package, its distribution path, or the mutable repository revision.

Attack Path

  1. An attacker compromises the npm package resolved by npx skills, its maintainer account, or another relevant distribution component.
  2. Alternatively, an attacker causes the mutable repository reference to resolve to an unauthorized revision.
  3. A user follows the documented installation or update command.
  4. npx retrieves and executes the externally supplied installer version with the user's local privileges.
  5. The compromised installer can run arbitrary code within those privileges or install altered skill content.

Impact Assessment

Successful exploi ...[truncated 444 chars]

Remediation
View remediation

Remediation Suggestions

  1. Pin the npm installer to a reviewed exact version, for example by using npx -y package-name@X.Y.Z ....
  2. Pin the installed repository to an immutable release tag or, preferably, a full reviewed commit SHA.
  3. Ensure update instructions use the same immutable and version-pinned approach as installation instructions.
  4. Publish checksums or signatures for release artifacts and instruct users to verify them before import.
  5. Consider using npm exec --package=package-name@X.Y.Z -- ... to make the selected installer version explicit.
  6. Review each new installer and skill release before changing the pinned versions.
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (10)

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The declared description is about a social media publishing skill using ReplyNodes, scheduling, cross-posting, and browser OAuth integrations with LinkedIn and X/Twitter. The supplied code does none of that. Instead, it is a package validation/security policy script for the skill repository itself. Its primary function is to enforce repository contents, exact SKILL.md metadata, secret scanning, shell safety rules, dependency hygiene, and release metadata consistency. Those are materially different behaviors and constitute undeclared capabilities unrelated to social media publishing or OAuth flows.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The declared description describes a social media publishing/scheduler integration with OAuth flows for LinkedIn and X/Twitter. The actual code is a minimal bash script whose notable behavior is executing a string as a shell command using eval. That is materially different from the stated purpose and introduces an undeclared capability—arbitrary or dynamic shell execution. There is no indication of approval workflows, ReplyNodes scheduling, cross-posting, auto-posting, OpenClaw integration, browser OAuth, or interaction with social media platforms. Therefore this is a clear description-behavior mismatch.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The declared description is about a social media publishing/scheduling skill with OAuth integrations for LinkedIn and X/Twitter. The supplied code does nothing related to publishing, scheduling, cross-posting, browser OAuth, or social media APIs. Instead, it is a repository/package test script that validates package structure and security-related constraints using fixtures. This is a materially different primary purpose and includes undeclared capabilities around package validation and security checks.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
82% confidence
Finding

The skill describes shell-relevant safety guidance but declares no explicit tool scope or permissions boundary. In agent environments, omission of allowed-tools/permissions can let the host expose broader execution capability than intended, increasing the chance that future prompts or workflow glue invoke shell access without a least-privilege constraint.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
85% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 81)May include surrounding context.

md
ReplyNodes can report connected channel health and explicit `publish` and `schedule` grants. A channel is supported for the requested action only when its returned capability includes that exact action: `publish` is not evidence of `schedule`, and a connected channel without either grant is unavailable for both. Availability depends on the user's external ReplyNodes account and connected provider accounts; this skill has no paywall, while ReplyNodes may be a paid service.

The skill cannot grant capabilities, reconnect expired providers, bypass approval, select another tenant, expose OAuth tokens, or publish to channels whose capability is absent. If a channel is expired, disabled, or disconnected, show the API's reason and ask the user to reconnect it in ReplyNodes.

## Errors

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The script invokes eval on a shell variable, which is an unsafe execution primitive because any future change that allows untrusted input into that variable can lead to arbitrary command execution. Even though the current value is the constant string 'true', the pattern is dangerous and the script provides no warning or restriction around shell execution.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

The README lists several invocation examples, including the phrase "publish everywhere," as a way to use the skill. That phrase is comparatively broad and could collide with ordinary conversation about publishing content widely, without clearly delimiting when the skill should or should not activate.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
97% confidence
Finding

The dependency is specified with a lower-bound range (>=1.2.3) instead of an exact version, which allows future installs to resolve to different, unreviewed releases. This weakens build reproducibility and can expose the project to supply-chain risk if a later version introduces malicious code, a compromise, or a breaking security regression.

Content

Scanner excerpt · tests/fixtures/unpinned-dependency/requirements.txt (reported line 1)May include surrounding context.

text
example-package>=1.2.3