Security audit

AI资讯速递

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent AI-news briefing helper; its main risk is optional chart widgets that load unpinned Chart.js from a CDN.

Reasonable to install for AI-news briefings. Prefer using SVG or a pinned/local Chart.js version for widgets, and ask for English or bilingual output if Chinese is not desired.

A.I.G

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:134
Finding: Unpinned Third-Party JavaScript Loaded into Visualization Widgets
Content: View full analysis
``` The surrounding example in `SKILL.md:132-136` instructs the agent to place this dependency in a rendered widget: ```html
Remediation: View remediation
/dist/chart.umd.min.js" integrity="" crossorigin="anonymous"> ``` The version and integrity value must be obtained from and verified against the selected release rather than copied from an untrusted source. 2. **Apply Subresource Integrity** Add a valid `integrity` attribute so the browser rejects content that does not match the audited artifact. Retain `crossorigin="anonymous"` where required for SRI validation. 3. **Prefer a locally bundled dependency** Store a reviewed Chart.js artifact in the trusted application distribution and load it locally. This removes runtime dependence on mutable third-party delivery infrastructure. 4. **Restrict widget capabilities** Run generated widgets in a sandboxed iframe without same-origin privileges unless strictly required. Deny access to host APIs, credentials, local storage, and sensitive DOM content. 5. **Enforce a restrictive Content Security Policy** Limit `script-src` to approved, immutable resources and disable unexpected outbound connections through restrictive `connect-src`, `img-src`, and related directives. 6. **Update all documented templates consistently** Correct the dependency reference in: - `SKILL.md:134` - `references/visualization_examples.md:11` - `references/visualization_examples.md:45` - `references/visualization_examples.md:106` 7. **Establish dependency maintenance controls** Document the approved version, review security advisories before upgrades, verify artifact hashes during updates, and avoid floating package versions in future examples. ]]>

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The manifest description lists the trigger phrase “帮我整理AI新闻等”, where the “等” leaves the activation scope open-ended and makes the trigger set ambiguous. This increases the chance of unintended invocation because users may use similar everyday wording without intending to call this specific skill.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: Line L139 states that chart titles and labels must use Chinese and match the briefing language. This is a natural-language locale policy constraint presented as a requirement, but the file does not offer a user choice or document a justified region-specific limitation.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 91% confidence
Finding: This markdown file presents all instructions and source descriptions exclusively in Chinese, including headings and annotations, with no indication that users may choose another language. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale constraint is clearly documented and justified.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The entire template and its headings are written only in Chinese, which implicitly constrains the skill to a specific language/locale without indicating user choice or opt-in. The policy for this audit flags language/locale restrictions unless the file offers a language choice or clearly documents a justified regional constraint.
Content: No source excerpt is available for this finding.