Back to skill

Security audit

kimi2.6free

Security checks for vulnerabilities and agentic risk

Overview

This skill is not clearly malicious, but it asks users to enable recurring authenticated automation that can change their forum account state and stores powerful secrets in predictable plaintext files.

Review carefully before installing. Avoid enabling the cron heartbeat unless you explicitly want recurring unattended forum actions under your account. Prefer environment variables or a secret manager over the documented credentials.json file, restrict any local credential file permissions, and use a narrowly scoped token if the service supports it. Treat the API key and node secret as account-control secrets.

Vulnerability Patterns
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • System PersistenceInstalls backdoors, hooks, services, or scheduled tasks that survive the run
  • 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
Findings (5)

T06 · System Persistence

Error
Location
HEARTBEAT-SETUP.md:30
Finding

Persistent Scheduled Agent Task Performs Unattended State-Changing Actions

Content
View full analysis
Remediation
View remediation

T02 · Agent Memory Poisoning

Error
Location
HEARTBEAT-SETUP.md:36
Finding

Scheduled Remote Interaction Is Written into Persistent Agent Memory

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
REGISTRATION.md:75
Finding

API Keys and Node Secrets Are Stored in Plaintext Without Permission Hardening

Content
View full analysis
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Warning
Location
index.js:6
Finding

Public Module Interface Returns Raw Credentials and Credential File Location

Content
View full analysis
loadCredentials(), maskSecret, ``` The returned object includes all loaded secrets: ```javascript const merged = { apiKey: envConfig.apiKey || fileConfig.apiKey || fileConfig.api_key, agentId: envConfig.agentId || fileConfig.agentId || fileConfig.agent_id, nodeSecret: envConfig.nodeSecret || fileConfig.nodeSecret || fileConfig.node_secret, agentName: envConfig.agentName || fileConfig.agentName || fileConfig.agent_name, apiBaseUrl: envConfig.apiBaseUrl || fileConfig.apiBaseUrl || fileConfig.api_base_url || 'https://www.singularity.mba', hubBaseUrl: envConfig.hubBaseUrl || fileConfig.hubBaseUrl || fileConfig.hub_base_url || 'https://www.singularity.mba', configPath: CONFIG_FILE, }; ``` ### Technical Analysis The package exports `getCredentials()` as part of its public interface. Calling it returns the raw API key, node secret, agent identity, endpoint configuration, and predictable credential file path. The forum API wrapper shown in this project only uses `config.apiKey` for authorization. Returning `nodeSecret` and the credential path to every module consumer is unnecessary for those operations and violates least-privilege design. It also makes accidental serialization or logging of the complete object more likely. ### Attack Path 1. Another plugin, package, or script running in the same Node.js process imports this module. 2. It invokes the publicly exported `getCredentials()` function. 3. The function reads environment variables or the credential file and returns the complete secret object. 4. T ...[truncated 478 chars]
Remediation
View remediation

other

Error
Location
lib/heartbeat.js:58
Finding

Heartbeat Automatically Applies Remote Genes and Upvotes Unreviewed Posts

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (31)

Credential Access

High
Category
Privilege Escalation
Confidence
97% confidence
Finding

The content names a fixed credentials file path and instructs users to place sensitive values there, including an API key and node secret. Publishing a standard secret location makes credential harvesting easier for malware, local attackers, or misconfigured tooling, and the example normalizes plaintext storage of reusable secrets.

Content

Scanner excerpt · REGISTRATION.md (reported line 76)May include surrounding context.

md
注册成功后**立即**保存以下信息:

**凭证文件路径:**
- Windows: `%APPDATA%\singularity\credentials.json`
- Linux/macOS: `~/.config/singularity/credentials.json`

**凭证内容(把注册返回的真实值填入):**

Credential Access

High
Category
Privilege Escalation
Confidence
97% confidence
Finding

This section continues the instruction to use a predictable local credentials.json file for storing sensitive authentication material. In the context of a registration guide, this is more dangerous because it operationalizes credential persistence without any safeguards, increasing the likelihood of theft and subsequent unauthorized API or agent access.

Content

Scanner excerpt · REGISTRATION.md (reported line 77)May include surrounding context.

凭证文件路径:

  • Windows: %APPDATA%\singularity\credentials.json
  • Linux/macOS: ~/.config/singularity/credentials.json

凭证内容(把注册返回的真实值填入):

json

Credential Access

High
Category
Privilege Escalation
Confidence
91% confidence
Finding

The skill discloses specific local paths and variable names for sensitive credentials, including API keys and node secrets, which indicates the skill is designed to locate and potentially consume secrets from the host environment. In an agent setting, referencing exact credential files materially increases the risk of unauthorized secret collection, exfiltration, or misuse if paired with code that reads those locations.

Content

Scanner excerpt · SKILL.md (reported line 69)May include surrounding context.

md
路径(按顺序读取):
1. 环境变量:`SINGULARITY_API_KEY`、`SINGULARITY_AGENT_ID`、`SINGULARITY_NODE_SECRET`
2. Windows:`%APPDATA%\singularity\credentials.json`
3. Linux/macOS:`~/.config/singularity/credentials.json`

## Forum API Base URL

Credential Access

High
Category
Privilege Escalation
Confidence
90% confidence
Finding

This line continues the disclosure of platform-specific credential file locations, reinforcing that the skill expects secrets to be present in predictable local files. That predictability lowers the barrier for hostile or over-privileged skill code to harvest credentials from the host, making the overall context more dangerous than a generic configuration example.

Content

Scanner excerpt · SKILL.md (reported line 70)May include surrounding context.

md
路径(按顺序读取):
1. 环境变量:`SINGULARITY_API_KEY`、`SINGULARITY_AGENT_ID`、`SINGULARITY_NODE_SECRET`
2. Windows:`%APPDATA%\singularity\credentials.json`
3. Linux/macOS:`~/.config/singularity/credentials.json`

## Forum API Base URL

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 · HEARTBEAT-SETUP.md (reported line 100)May include surrounding context.

md
* 
 * 按以下顺序读取凭证:
 * 1. 环境变量
 * 2. Windows: %APPDATA%\singularity\credentials.json
 * 3. Linux/macOS: ~/.config/singularity/credentials.json
 */

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 · HEARTBEAT-SETUP.md (reported line 105)May include surrounding context.

md
* 
 * 按以下顺序读取凭证:
 * 1. 环境变量
 * 2. Windows: %APPDATA%\singularity\credentials.json
 * 3. Linux/macOS: ~/.config/singularity/credentials.json
 */

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 · lib/config.js (reported line 7)May include surrounding context.

js
* 
 * 按以下顺序读取凭证:
 * 1. 环境变量
 * 2. Windows: %APPDATA%\singularity\credentials.json
 * 3. Linux/macOS: ~/.config/singularity/credentials.json
 */

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 · lib/config.js (reported line 8)May include surrounding context.

js
* 
 * 按以下顺序读取凭证:
 * 1. 环境变量
 * 2. Windows: %APPDATA%\singularity\credentials.json
 * 3. Linux/macOS: ~/.config/singularity/credentials.json
 */

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 · lib/config.js (reported line 18)May include surrounding context.

js
* 
 * 按以下顺序读取凭证:
 * 1. 环境变量
 * 2. Windows: %APPDATA%\singularity\credentials.json
 * 3. Linux/macOS: ~/.config/singularity/credentials.json
 */

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 · lib/heartbeat.js (reported line 32)May include surrounding context.

js
* 
 * 按以下顺序读取凭证:
 * 1. 环境变量
 * 2. Windows: %APPDATA%\singularity\credentials.json
 * 3. Linux/macOS: ~/.config/singularity/credentials.json
 */

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 · EXPERIENCE-CARD.md (reported line 87)May include surrounding context.

curl:

bash
curl -X POST https://www.singularity.mba/api/proxy/v1/chat/completions \
  -H "Authorization: Bearer <your_api_key>" \
  -H "Content-Type: application/json" \
  -d '{

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 · EXPERIENCE-CARD.md (reported line 99)May include surrounding context.

Node.js:

javascript
const response = await fetch('https://www.singularity.mba/api/proxy/v1/chat/completions', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer <your_api_key>',

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The documentation directs users to install a recurring cron job that automatically performs account actions such as fetching/applying genes, sending heartbeats, upvoting, commenting, and writing local memory files. Because it does not clearly warn that this will modify remote account state and create local artifacts on a schedule, users may enable ongoing autonomous behavior without informed consent, increasing the risk of unintended account activity, spammy behavior, and filesystem side effects.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The credential section instructs users to place API keys and node secrets in a local JSON file but provides no warning about the sensitivity of those secrets, storage permissions, or the consequences of compromise. This can lead to insecure handling of long-lived credentials that enable authenticated API access and node impersonation if exposed through logs, backups, or weak filesystem permissions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

This markdown file contains user-facing instructions only in Chinese, which can amount to a language/locale policy issue when no user opt-in or alternative language option is provided. The content does not state that it is intentionally limited to a Chinese-speaking audience or region-specific use case.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The document instructs users to store a forum API key directly in a local JSON config and transmit it to remote service endpoints, but provides no guidance on protecting the secret at rest or avoiding exposure through shell history, shared config files, backups, or logs. In this skill context, that omission is meaningful because the API key appears to authorize forum access and possibly model-provider access, so credential leakage could enable account misuse, data access, or impersonation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The guide explicitly instructs users to store long-lived secrets locally, including an API key and node secret, but provides no warning about filesystem permissions, encryption, or secret-handling practices. This increases the chance of credential exposure through local compromise, backups, logs, screenshots, or accidental commits, enabling unauthorized access to the associated account and APIs.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill explicitly documents where credentials are sourced from, including environment variables and local credential files, but provides no warning about sensitive handling, storage protection, or avoiding disclosure to untrusted tools. In an agent-skill context, this can normalize credential access and increase the chance that downstream code or users expose API keys and node secrets to the skill or related automation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This module exposes many account-affecting operations through a thin wrapper that automatically loads stored credentials and immediately performs actions such as applying genes, marking notifications read, posting comments, upvoting posts, exchanging cards, and sending heartbeats. In an agent-skill context, this creates a dangerous capability surface because callers can trigger state-changing actions without explicit user confirmation, scope checks, or per-action consent, increasing the risk of unintended or unauthorized account activity if the skill is invoked improperly.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

This code builds Authorization headers from config.apiKey and uses them across multiple fetch calls, including POST requests that send notification, comment, heartbeat, and other user-related data to a remote service. While the endpoints are commented, there is no user-facing prompt, log, or warning in this file disclosing that the skill will send authenticated data to an external API.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The file presents itself as a periodic 'heartbeat' script, but it also performs multiple state-changing actions: marking notifications read, applying genes, sending heartbeats, and upvoting posts. This mismatch is dangerous because operators may run it expecting passive status reporting while it silently modifies account state and performs social actions on their behalf.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% 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.

Content

Scanner excerpt · lib/heartbeat.js (reported line 32)May include surrounding context.

js
const config = loadCredentials();

  if (!config.apiKey) {
    log('error', 'No API key found. Set SINGULARITY_API_KEY env or create ~/.config/singularity/credentials.json');
    process.exit(1);
  }

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The script can automatically mark notifications as read, which alters user-visible account state and may hide important unread items or audit signals. Although gated by a CLI flag, there is no confirmation prompt, preview, or warning at runtime, so the action can be triggered in automation with little user awareness.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
73% confidence
Finding

All user-facing instructions in this file are presented in Chinese, and the document does not offer an alternate language or state that the skill is intended only for a Chinese-speaking audience. This can violate a language/locale policy when users are not given a choice or informed of the locale constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

A language or locale policy issue can arise when a skill forces a specific language without user opt-in. This file presents all instructions exclusively in Chinese and does not indicate that users may choose another language or that the scope is intentionally Chinese-only.

Content

No source excerpt is available for this finding.