Back to skill

Security audit

Meyo Community

Security checks for vulnerabilities and agentic risk

Overview

This skill appears purpose-built for Meyo community actions, but it uses local credentials to make authenticated network requests and has an unsafe credential-file parsing bug that could allow local code execution if the environment is influenced.

Review this before installing. Use it only if you intend an agent to access your Meyo account, send requests to meyo123.com, and potentially publish posts under your account. Keep the credentials file protected, avoid untrusted MEYO_CRED_FILE values or inherited environments, and require a manual preview/confirmation workflow before running the post command.

Vulnerability Patterns
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T09 · Insecure Skill Coding Practices

Error
Location
scripts/meyo.sh:8
Finding

Arbitrary Python Code Execution Through an Unsafely Interpolated Credential Path

Content
View full analysis
/dev/null } ``` ### Technical Analysis The `MEYO_CRED_FILE` environment variable controls `CRED_FILE`. The value is then inserted directly into source code passed to `python3 -c`, inside a nominally single-quoted Python string. The shell's double quotes do not make this safe: they allow `$CRED_FILE` to be expanded, after which the expanded value becomes part of the Python program. A path containing quotes, parentheses, semicolons, comments, or other Python syntax can terminate the intended string and inject additional statements. The later file-existence check does not prevent this issue. An attacker with control over both the environment and writable filesystem locations can create a file whose literal name matches the crafted `MEYO_CRED_FILE` value so it passes `[ -f "$CRED_FILE" ]`. The injected Python expression can refer to a separate attacker-created valid JSON file before executing an appended payload. The `$1` field name is also interpolated into Python source. Its current callers use the fixed value `api_key`, so it is not presently attacker-controlled through the exposed command interface. Nevertheless, both values should be passed as data rather than embedded into executable source. Credential-file access itself is necessary for the advertised authenticated community operations and is limited to the `api_key` property. The security issue is not the use of a dedicated credential file, but the unsafe method used to parse an environment-selectable path. ### Attack Path 1. An attacker gains in ...[truncated 1768 chars]
Remediation
View remediation
/dev/null`; return a sanitized error while preserving useful diagnostics through an appropriate protected logging channel. 6. Keep the API key out of command-line arguments and diagnostic output. 7. Add tests using paths containing quotes, whitespace, newlines, shell metacharacters, and Python syntax to verify that all path values remain inert data. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (12)

Tp4

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding

The documented purpose does not fully match behavior: it references growth diary functionality without an implemented interface and understates heartbeat data collection by not clearly disclosing recommendations and announcements. This reduces informed consent and can cause the agent to perform or expose actions/data outside what a user reasonably expects.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
91% confidence
Finding

The skill explicitly depends on a local credentials file and states that all operations read authentication information from it. Access to stored credentials is not inherently malicious, but in an agent skill that can run shell commands and interact with an external service, it creates a high-value trust boundary: accidental invocation or misuse could perform authenticated actions or expose account data.

Content

Scanner excerpt · SKILL.md (reported line 12)May include surrounding context.

md
## 前置条件

凭证文件:`~/.openclaw/meyo/credentials.json`(本地配置,不随 skill 分发)。所有操作从该文件读取认证信息。

## 核心操作

Credential Access

High
Category
Privilege Escalation
Confidence
78% confidence
Finding

The documentation explicitly points agents or operators to a local credential file at a fixed path, which facilitates credential discovery and use for authenticated API actions. In an agent-skill context, this is more dangerous because the skill is designed to perform community operations and could encourage automated access to bearer tokens stored on disk, increasing the chance of unintended use, leakage, or abuse if surrounding controls are weak.

Content

Scanner excerpt · references/api-reference.md (reported line 6)May include surrounding context.

md
## 基础信息

- Base URL: `https://www.meyo123.com/api/v1`
- 凭证: `~/.openclaw/meyo/credentials.json`
- 统一脚本: `scripts/meyo.sh`

## 认证

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 · scripts/meyo.sh (reported line 8)May include surrounding context.

sh
set -euo pipefail

CRED_FILE="${MEYO_CRED_FILE:-$HOME/.openclaw/meyo/credentials.json}"
BASE_URL="https://www.meyo123.com/api/v1"

# ── 认证 ──────────────────────────────────────────────

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
89% confidence
Finding

The skill invokes a shell script but does not declare any tool scope or allowed-tools boundary. That makes the runtime authority implicit rather than auditable, increasing the chance the agent can execute shell actions unexpectedly or with broader access than users realize.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The trigger phrases are broad enough that the skill may activate for generic requests about community activity, posting, or diaries. Unintended invocation is risky here because the skill can use stored credentials and perform external actions such as posting content or querying private account-related interactions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill describes posting and interaction features without warning that user-provided content will be sent to an external service using locally stored credentials. This can lead to silent publication, account actions on behalf of the user, and privacy surprises, especially when prompts contain sensitive or draft content.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
88% confidence
Finding

The script transmits data and authentication headers to an external domain via curl, including searches, heartbeat data, and post payloads. In the context of an agent skill, any external transmission is security-relevant because it can send user-provided data or metadata off-host without strong visibility or consent controls.

Content

Scanner excerpt · scripts/meyo.sh (reported line 35)May include surrounding context.

sh
)

api_get() {
  curl -s "${BASE_URL}${1}" "${AUTH_HEADERS[@]}"
}

api_post() {

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The script can publish arbitrary user-supplied title/content/tag data to a live remote community endpoint with no confirmation, dry-run mode, or explicit disclosure at execution time. In an agent setting, this increases the risk of unintended posting, prompt-driven abuse, spam, or accidental disclosure of sensitive information because the action is state-changing and externally visible.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
72% confidence
Finding

The file's natural-language instructions are entirely in Chinese and do not appear to offer an alternative language or explain a justified locale restriction. Under the policy rule, forcing a specific language without user opt-in can be a natural-language policy violation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

This markdown file documents use of a local credential store at ~/.openclaw/meyo/credentials.json and an Authorization: Bearer <api_key> header, but it does not include any warning about protecting secrets, avoiding accidental disclosure, or ensuring secure handling. For markdown files, missing warnings about privacy- or security-relevant behaviors can qualify as a missing user warning.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The script loads an API key from ~/.openclaw/meyo/credentials.json and uses it in an Authorization header, but there is no explicit notice in the script output or inline disclosure that local credentials will be accessed for remote authentication. Access to credentials is a sensitive operation and should be made visible to users unless clearly disclosed elsewhere.

Content

No source excerpt is available for this finding.