Back to skill

Security audit

用天气可视化你的心情!基于 AI 情绪分析,将文字转化为天气图标 + 治愈文案。

Security checks for vulnerabilities and agentic risk

Overview

This mood-analysis skill appears purpose-built, but it relies on unreviewed globally installed code and broad triggers that could send sensitive mood text to an external service unexpectedly.

Review this carefully before installing. Only install if you are comfortable with a global npm package running on your machine, using a DeepSeek API key, and potentially sending mood-related text to DeepSeek. Prefer a pinned, locally installed, fully reviewed CLI implementation, restrict the API-key file permissions, and avoid submitting highly sensitive personal or mental-health details until privacy and retention behavior are clearer.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (2)

T08 · Insecure Dependencies

Error
Location
scripts/install.sh:24
Finding

Unpinned Global npm Package Is Installed and Immediately Executed

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/install.sh:37
Finding

API-Key Configuration File Is Created Without Restrictive Permissions

Content
View full analysis
"$ENV_FILE" echo "DEEPSEEK_API_KEY=sk-your-api-key-here" >> "$ENV_FILE" fi ``` ### Technical Analysis The installer creates `~/.mood-weather-cli.env` using ordinary shell redirection without setting a restrictive `umask` or explicitly applying mode `0600`. The resulting permissions depend on the user's ambient `umask`. On systems configured with permissive defaults, the file may be readable by other local users or processes. The documentation instructs the user to replace the placeholder with a real DeepSeek API key, turning this file into a plaintext secret store. The conditional check also leaves the permissions of an existing file unchanged, even if that file is already excessively permissive. ### Attack Path 1. A user runs the installation script with a permissive `umask`. 2. The script creates `~/.mood-weather-cli.env` without enforcing owner-only access. 3. The user edits the file and inserts a valid DeepSeek API key as instructed. 4. Another local account or process with filesystem access reads the configuration file. 5. The attacker uses the recovered key to make unauthorized API requests or exhaust the victim's API quota. ### Impact Assessment Exploitation can disclose the DeepSeek API key to local users or processes that can read the file. The exposed key could be used to: - Submit unauthorized requests under the victim's account. - Consume paid API quota or trigger service limits. - Access capabilities granted to that API credential. - Correlate or interfere with the victim's use of the service. This issue does not independently grant system-level privileges; its scope is limited to credentials an ...[truncated 53 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (24)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill description says it analyzes mood via a CLI tool but does not warn that the text may be sent to an external API service using DEEPSEEK_API_KEY. For a mood-analysis skill, the input is likely to contain sensitive emotional or mental-health-adjacent data, so omission of this disclosure undermines informed consent and increases privacy risk.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The changelog states the skill supports natural-language triggering via broad keywords such as “mood”, “心情”, and “情绪”, which are generic terms likely to appear in ordinary conversation. Overly broad activation criteria can cause unintended skill invocation, sending user text to the mood-analysis CLI or downstream LLM when the user did not explicitly request that action, creating privacy and routing risks.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The README states that mood text is analyzed by an external AI service (DeepSeek) but does not clearly warn users that their potentially sensitive emotional disclosures will be transmitted off-device. Because mood and mental-state text can contain highly personal information, the lack of an explicit privacy/data-transfer notice may cause users to disclose sensitive content without informed consent.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The README advertises automatic saving of analysis history and future emotion statistics/reporting, but it does not warn users that sensitive mood-analysis results may be stored locally. Local persistence of emotional data can expose private information to other local users, backups, malware, or accidental disclosure if retention behavior is not clearly explained.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger list includes broad everyday phrases like '心情怎么样' and '心情好吗' that can easily appear in normal conversation, causing the skill to activate when the user did not clearly intend to invoke an external mood-analysis tool. In this skill, unintended activation is more dangerous because the analyzed text may be sent to a third-party API handling sensitive emotional content.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

Keyword-based activation on generic phrases such as '分析情绪', '今天心情', and '心情好/不好' is overly permissive and can match ordinary chat content. Because the skill is designed to process free-form user text and may forward it to DeepSeek, accidental activation can expose sensitive personal content without a deliberate user action.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The planned history, statistics, report, and batch-analysis features describe retaining emotional inputs and analysis results over time without defining retention limits, minimization, access controls, or user consent boundaries. Emotional self-expression is sensitive data, so persistent storage materially increases privacy harm if the host, logs, or exported reports are accessed by others.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The best-practice guidance recommends logging every invocation result, which in this skill can include highly sensitive user emotional text, inferred mood labels, and confidence values. Centralized logs often have broader retention and wider operator access than application data, so this creates avoidable secondary exposure of sensitive information.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The script performs npm install -g mood-weather-cli, which installs executable code system-wide and places a new command on the user's PATH. Global package installation increases blast radius if the package is compromised, typosquatted, or later updated maliciously, and it exceeds the least-privilege model for a narrowly scoped sentiment-analysis skill.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The install script introduces and operationalizes a remote DeepSeek API dependency by creating an API-key config and running a health check, while the skill description only presents local mood analysis via a CLI. This mismatch matters because user text may be sent to a third-party service without clear disclosure, expanding the data exposure and trust boundary beyond what users would reasonably expect.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The script creates a plaintext environment file intended to hold a live API key but does not set restrictive permissions or clearly warn that sensitive credentials will be stored unencrypted on disk. On multi-user or backed-up systems, this can lead to accidental credential exposure through permissive file modes, logs, sync tools, or user mishandling.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

Several trigger patterns are broad natural-language phrases such as "my mood", "心情怎么样", and "how am I feeling", which can easily appear in ordinary conversation without the user intending to invoke a CLI-backed skill. Because the skill uses extractArg: true and executes the mood command, ambiguous activation expands the attack surface for accidental or adversarial prompt routing.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The help trigger includes the generic phrase "帮助", which is a common everyday word and can invoke this skill unintentionally in many unrelated conversations. In an agent environment, broad triggers increase the chance of accidental command execution and misrouting user intent, especially because this skill launches executable commands for help output.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

Most of the skill's user-facing instructions and examples are in Chinese, while the document does not state that the skill is intended only for Chinese-speaking users or offer an explicit language option. This can amount to an implicit language policy restriction without user opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The package description is written entirely in Chinese, which can constitute a language/locale policy issue when no alternative language option or opt-in is provided. In this manifest, there is no indication that the skill is intentionally region-specific or that users can choose another language.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
89% confidence
Finding

Using a caret range for node-fetch allows future installs to resolve to different patch/minor versions, which weakens supply-chain reproducibility and makes security posture harder to verify. In a CLI skill that may process user input and perform network requests, dependency drift can unexpectedly introduce vulnerable or malicious code.

Content

Scanner excerpt · package.json (reported line 36)May include surrounding context.

json
"node": ">=14.0.0"
  },
  "dependencies": {
    "node-fetch": "^2.6.9",
    "commander": "^9.4.0",
    "dotenv": "^16.0.3",
    "axios": "^1.4.0"

Unverifiable Dependency: node-fetch has 3 known advisory(ies) (CVE-2022-0235 (node-fetch forwards secure headers to untrusted sites); CVE-2022-2596 (node-fetch Inefficient Regular Expression Complexity ); CVE-2020-15168 (The `size` option isn't honored after following a redirect in node-fetch)), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
86% confidence
Finding

The manifest references node-fetch with a non-exact version while known advisories exist for some node-fetch releases, so the actual installed package may be affected and cannot be reliably verified from this file alone. Because this CLI likely performs outbound HTTP requests, issues involving header forwarding or redirect handling could expose tokens or sensitive request metadata in certain configurations.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
88% confidence
Finding

Using an unpinned commander version reduces build determinism and increases supply-chain risk because different environments may install different releases over time. While commander itself is a common CLI library, version ranges still make it harder to attest what code is actually executed.

Content

Scanner excerpt · package.json (reported line 37)May include surrounding context.

json
},
  "dependencies": {
    "node-fetch": "^2.6.9",
    "commander": "^9.4.0",
    "dotenv": "^16.0.3",
    "axios": "^1.4.0"
  }

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
88% confidence
Finding

An unpinned dotenv dependency permits dependency drift, which can undermine reproducibility and make incident response or vulnerability verification harder. Because dotenv influences runtime configuration loading, unexpected package changes could affect secret handling behavior.

Content

Scanner excerpt · package.json (reported line 38)May include surrounding context.

json
"dependencies": {
    "node-fetch": "^2.6.9",
    "commander": "^9.4.0",
    "dotenv": "^16.0.3",
    "axios": "^1.4.0"
  }
}

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
92% confidence
Finding

Using a caret range for axios is a supply-chain risk and is more significant here because axios is a network-facing HTTP client commonly implicated in SSRF, redirect, and request-handling issues. In a sentiment-analysis CLI that likely calls external APIs, an unexpected resolved version could materially affect network security behavior.

Content

Scanner excerpt · package.json (reported line 39)May include surrounding context.

json
"node-fetch": "^2.6.9",
    "commander": "^9.4.0",
    "dotenv": "^16.0.3",
    "axios": "^1.4.0"
  }
}

Unverifiable Dependency: axios has 16 known advisory(ies) (CVE-2026-44494 (axios Vulnerable to Full Man-in-the-Middle via Prototype Pollution Gadget in `co); CVE-2026-44495 (axios Vulnerable to Credential Theft and Response Hijacking via Prototype Pollut); CVE-2025-62718 (Axios has a NO_PROXY Hostname Normalization Bypass that Leads to SSRF) +13 more), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
91% confidence
Finding

Axios has a history of multiple security advisories, and the manifest does not pin an exact version, so the deployed package may resolve to an affected release without easy verification. This is more concerning in a CLI skill that likely sends user text to remote services, where SSRF, proxy bypass, or request-manipulation flaws could impact confidentiality or network boundary controls.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

Comments, status messages, and usage text are presented only in Chinese, which imposes a specific language on users without opt-in or alternative locale support. This is a natural-language policy concern because the skill does not offer any language selection or document a justified locale restriction.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

The manifest presents the skill description in Chinese, and the output prefix is also fixed in Chinese, while the trigger set includes English phrases. This suggests the skill may respond in a specific language without offering user opt-in or documenting a locale-specific scope.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

The output prefix and several help-related strings are hardcoded in Chinese, but the manifest also accepts English triggers like "analyze mood" and "how am I feeling". This creates a natural-language policy concern because users invoking the skill in English are not given a language choice or notice of the forced locale.

Content

No source excerpt is available for this finding.