News Source Manager (信息源管理器)

• Outcome: Pass
• Latest audit: Sep 15, 2026 · 1:21 AM UTC
• Version: 1.0.2

Overview

This skill transparently manages local news preference settings and does not show deceptive, destructive, or unrelated behavior.

Before installing, be aware that your selected news topics and preferred sources will be saved locally in ~/.openclaw/workspace/memory/news-sources.json and may reveal interests such as finance, health, policy, or work focus on shared or synced machines. Review or delete that file if you no longer want the preferences retained.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• YARA Signatures: Malware Match, Webshell Match, Cryptominer Match
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (9)

Natural-Language Policy Violations

Medium

• Confidence: 90% confidence
• Finding: The documented first-use interaction is entirely in Chinese and is presented as what the user will be asked when the skill runs. Because the README does not say the user can choose another language, this creates a natural-language policy concern around forcing a specific language without user opt-in.

Missing User Warnings

Medium

• Confidence: 86% confidence
• Finding: The README explicitly states that user news preferences are stored on disk in a predictable path, but it does not warn users about persistence, local exposure, or privacy implications. Preference data can reveal interests, profession, investment focus, or health/policy concerns, and undocumented retention increases the risk of unintentional disclosure on shared or synced systems.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The invocation description allows activation for a broad, common user intent ('wants to customize news categories') rather than requiring a clear command scoped to this skill. That can cause the skill to trigger in unrelated conversations and perform reads or writes to the user's persistent configuration without sufficiently explicit intent.

Natural-Language Policy Violations

Medium

• Confidence: 81% confidence
• Finding: The skill title and purpose force bilingual or Chinese-language presentation by default, and the operational examples throughout the file continue this mixed-language pattern without stating that the user can choose their preferred language. This can violate a language/locale policy where user-facing language should be selected by user preference or explicit opt-in.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The modify-sources trigger uses ordinary conversational phrasing like complaints about source quality, which may be mentioned hypothetically or as discussion rather than a direct command. This increases the risk of unintended configuration changes to a persistent file if the agent interprets ambient language as authorization to update settings.

YARA rule 'network_reconnaissance': Network reconnaissance and scanning patterns [hacktools]

Medium

• Category: YARA Match

Content

uctor, robotics, AI safety | TechCrunch, MIT Tech Review, The Verge |
| **Business Strategy** | M&A, disruption, SaaS, digital transformation, startup | McKinsey, HBR, WSJ, Bloomberg |
| **Finance/Crypto** | VC, IPO, bitcoin, DeFi, macroeconomics, fintech | Bloomberg, CoinDesk, FT, Reuters |
| **Health/Bio** | biotech, clinical trials, CRISPR, digital health, medical AI | STAT News, Nature, NEJM, Fierce Biotech |
| **Energy/Climate** | renewable, battery, EV, carbon capture, hydrogen, nuclear | Canary Media, CleanTechnica, Carbon Brief |
| **Policy/Regulation** | AI regulation, antitrust, GDPR, AI Act, cybersecurity | Politico, Bloomberg Gov, Lawfare |
| **Product Design** | UX, UI, design system, accessibility, Figma, user research | Nielsen Norman, Smashing Magazine |

**keywords 角色**: insight-radar 搜索后用 keywords 评估覆盖度，发现盲区自动补搜。不再作为搜索词。

• Confidence: 65% confidence
• Finding: YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Vague Triggers

Medium

• Confidence: 90% confidence
• Finding: The Finance/Crypto keyword list includes many generic terms such as "finance," "investment," "trading," "currency," "inflation," "law," and similar broad concepts that can appear in ordinary conversation. In a skill-selection or routing context, this can cause unintended activation, misclassification of user intent, and downstream handling of queries under the wrong policy or data path.

YARA rule 'network_reconnaissance': Network reconnaissance and scanning patterns [hacktools]

Medium

• Category: YARA Match

Content

"name": "Cell",
        "url": "cell.com",
        "priority": 2
      },
      {
        "name": "The Lancet",
        "url": "thelancet.com",
        "priority": 1
      },
      {
        "name": "NEJM",
        "url": "nejm.org",
        "priority": 1
      },
      {
        "name": "MedTech Dive",
        "url": "medtechdive.com",
        "priority": 2
      },
      {
        "name": "Fierce Biotech",
        "url": "fiercebiotech.com",
        "priority": 2
      },
      {
        "name": "Endpoints News",
        "url": "endpts.com",
        "priority": 2
      },
      {
        "name": "BioPharma Dive",
        "url": "biopharmadive.com",
        "priority": 2
      }
    ]
  },
  "Energy/Climate": {
    "keywords": [
      "renewable energy",
      "clean energy",
      "green energy",
      "solar energy",
      "solar panel",
      "photovoltaic",
      "PV",
      "wind energy",
      "wind turbine",
      "offshore wind",
      "onshore wind",
      "nuclear energ

• Confidence: 65% confidence
• Finding: YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The Policy/Regulation template contains broad triggers like "bill," "law," "Congress," "Senate," "House," and "surveillance," which can match many non-policy conversations. Because this skill manages news preferences and may be invoked by keyword-based intent detection, these unspecific terms increase the risk of accidental activation and incorrect preference or source management actions.

Static analysis

No suspicious patterns detected.