Back to skill

Security audit

solon-development-skill

Security checks for vulnerabilities and agentic risk

Overview

This Solon developer-reference skill is mostly transparent documentation, but it should go to Review because its security examples include copyable insecure authentication, CORS, and secret-handling patterns.

Review generated security-sensitive code before using it. Do not copy the auth, CORS, vault, logging, or agent-permission snippets directly into production; require real token validation, deny-by-default authorization, trusted-origin allowlists, unique externally managed secrets, retention controls, and least-privilege tool permissions.

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 (4)

T09 · Insecure Skill Coding Practices

Error
Location
references/security.md:97
Finding

Fail-Open Authentication and Authorization Processor

Content
View full analysis
0; } @Override public boolean verifyPath(String path, String method) { // Verify whether the user can access the path return true; } @Override public boolean verifyPermissions(String[] permissions, Logical logical) { // Triggered when verifyLogined returns true return true; } @Override public boolean verifyRoles(String[] roles, Logical logical) { // Triggered when verifyLogined returns true return true; } } ``` ### Technical Analysis The documented `AuthProcessor` implementation unconditionally approves IP-address, path, permission, and role checks. These methods form the authorization decisions used by the surrounding `AuthAdapter` example, but the documentation does not clearly identify them as insecure placeholders. Consequently, any application that copies this implementation may appear to have authorization controls while all meaningful policy checks fail open. A user who only satisfies—or bypasses elsewhere—the minimal logged-in condition can access paths and operations intended for specific roles or permissions. ### Attack Path 1. A developer copies the documented `AuthProcessorImpl` into an application. 2. The application registers it as the processor for its `AuthAdapter`. 3. An attacker sends a request to an endpoint protected by an IP, path, role, or permission rule. 4. The framework invokes the c ...[truncated 676 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
references/api_annotations.md:356
Finding

Authorization Filter Accepts Arbitrary Non-Null Tokens

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
references/security.md:10
Finding

Wildcard CORS Configuration Is Presented as the Default Pattern

Content
View full analysis
{ // Global handling using filter mode app.filter(-1, new CrossFilter().allowedOrigins("*")); // Handling for a path segment app.filter(new CrossFilter().pathPatterns("/user/**").allowedOrigins("*")); // Router interceptor mode app.routerInterceptor(-1, new CrossInterceptor().allowedOrigins("*")); }); ``` ### Technical Analysis All three documented approaches authorize every web origin through `origins = "*"` or `allowedOrigins("*")`. The examples include controller-wide, inherited, path-specific, and global configurations without demonstrating a trusted-origin allowlist or warning readers against using wildcard policies for sensitive APIs. CORS does not itself authenticate an attacker, but it controls whether browser scripts from attacker-controlled origins can access API responses. A wildcard policy unnecessarily expands the set of origins allowed to interact with the service and can amplify weaknesses in token storage, browser authentication, or other request controls. ### Attack Path 1. A developer copies one of the wildcard CORS examples into a deployed API. 2. A victim visits an attacker-controlled website. 3. The website’s JavaScript issues a cross-origin request to the API. 4. The API responds with a CORS policy authorizing the attacker’s origin through the wildcard configuration. 5. Where the browser and ...[truncated 741 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
references/security.md:174
Finding

Fixed Vault Key and Plaintext Credentials in Copyable Configuration

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (29)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 35)May include surrounding context.

md
| 注解对照 / IoC 容器 / 配置系统 / 插件SPI / SnEL表达式 / Spring对比 | `references/core_concepts.md` | `@Inject`, `@Configuration`, `app.yml`, `SnEL`, `SpiLoader`, `Spring` |

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 44)May include surrounding context.

md
| SSE 服务端推送 / Reactive 响应式 / I18n 国际化 | `references/web_advanced.md` | `SseEmitter`, `Flux`, `Mono`, `solon-web-sse`, `solon-web-rx`, `I18nUtil` |

Missing User Warnings

High
Category
Not specified by scanner
Confidence
94% confidence
Finding

This section documents agent skills and harness capabilities that include bash execution, file read/write/edit, web access, MCP integration, and dynamic subagent creation, but it presents them with minimal safety framing. In an agent-development skill, this is more dangerous because readers are likely to operationalize these patterns directly, potentially enabling destructive commands, unauthorized file changes, or broad tool abuse in real systems.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

This markdown file documents multiple behaviors that can affect privacy and user data, including remote chat/model API calls, reranking over external endpoints, web search, and loading local files into RAG pipelines. The description presents these capabilities as usage examples but does not include warnings about sending prompts, documents, or metadata to third-party services or processing local files.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 · references/ai_development.md (reported line 154)May include surrounding context.

Builder 原始构建:

java
RerankingModel rerankingModel = RerankingModel.of("https://api.moark.com/v1/rerank")
    .apiKey("***").model("bge-reranker-v2-m3").build();

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 · references/ai_development.md (reported line 162)May include surrounding context.

Builder 原始构建:

java
RerankingModel rerankingModel = RerankingModel.of("https://api.moark.com/v1/rerank")
    .apiKey("***").model("bge-reranker-v2-m3").build();

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 · references/ai_development.md (reported line 339)May include surrounding context.

Builder 原始构建:

java
RerankingModel rerankingModel = RerankingModel.of("https://api.moark.com/v1/rerank")
    .apiKey("***").model("bge-reranker-v2-m3").build();

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The example stores chat sessions in memory and attaches the sessionId to metadata returned through the stream, which affects user privacy and data retention expectations. The markdown does not disclose that conversations may persist across requests or that identifiers are propagated back to clients.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
75% confidence
Finding

Docker image references without a specific tag (:latest is implicit) or digest (@sha256:...) can be silently replaced by a malicious image.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill content is presented entirely in Chinese, beginning with the title, with no indication that language selection is optional or that the skill is intentionally region-specific. The policy requires flagging language or locale constraints when a specific language is imposed without user opt-in or clear justification.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The introductory text is written in Chinese and frames the document's applicable usage entirely in that language, while the file does not offer an alternative language option or state that the skill is intentionally region-specific. This can violate language/locale policy when users are not given opt-in or choice.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

This markdown file presents all instructional content in Chinese, including headings, explanations, and examples, but does not indicate that the user can choose another language. Under the policy for natural-language issues, forcing a specific language without user opt-in is a language/locale policy violation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This markdown file documents enabling file and cloud appenders and later shows persistent database insertion for logs, but it does not warn that logs may contain sensitive user, system, or credential-related information. Because the file is a markdown skill description, SQP-2 applies when potentially privacy- or integrity-affecting behavior is described without user-facing caution.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The section on a high-performance persistent appender explicitly shows batching log events into storage via logService.insertList(list), which can retain potentially sensitive operational or user data. The markdown does not disclose this privacy and retention impact, so it lacks the warning expected for markdown documentation describing data-affecting behavior.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 · references/remoting.md (reported line 198)May include surrounding context.

md
| 字段 | 说明 | 示例 |
|---|---|---|
| url | 完整的 url 地址 | `http://api.water.org/cfg/get/` |
| group | 服务组 | `water` |
| name | 服务名或负载均衡组件名(配合发现服务使用) | `waterapi` |
| path | 路径 | `/cfg/get/` |

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

This markdown file applies to SQP-2, and it includes multiple examples of outbound HTTP/RPC communication plus explicit token/header injection, such as headers = "TOKEN=xxx" and filters adding Token headers. The documentation does not include any warning about transmitting user or system data to remote services or handling secrets in headers, which is relevant to privacy and system integrity for a remoting skill description.

Content

No source excerpt is available for this finding.

Unsafe Defaults

Medium
Category
Tool Misuse
Confidence
90% confidence
Finding

The duplicated unsafe-defaults finding is substantiated by the same overall issue: the file normalizes insecure patterns in a document specifically about security. Because the skill context is developer guidance for Solon applications, unsafe snippets here are more dangerous than in abstract discussion since they are likely to be copied into live services.

Content

Scanner excerpt · references/security.md (reported line 1)May include surrounding context.

md
# Security — 安全(认证/鉴权/CORS/加密)

> 适用场景:跨域处理、用户认证、路径鉴权、角色权限控制。

Unsafe Defaults

Medium
Category
Tool Misuse
Confidence
90% confidence
Finding

The duplicated unsafe-defaults finding is substantiated by the same overall issue: the file normalizes insecure patterns in a document specifically about security. Because the skill context is developer guidance for Solon applications, unsafe snippets here are more dangerous than in abstract discussion since they are likely to be copied into live services.

Content

Scanner excerpt · references/security.md (reported line 1)May include surrounding context.

md
# Security — 安全(认证/鉴权/CORS/加密)

> 适用场景:跨域处理、用户认证、路径鉴权、角色权限控制。

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The documentation repeatedly demonstrates allowedOrigins("*") and @CrossOrigin(origins = "*") as normal usage without any warning that wildcard CORS should only be used for public, non-sensitive resources. In a security reference file, readers are likely to copy these examples into production, which can unintentionally expose authenticated APIs or sensitive responses to cross-origin web pages depending on credential and browser behavior.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The file includes a concrete vault password value in configuration examples without prominently stating that it is a placeholder and must never be reused. Even though it appears illustrative, publishing secret-like values in docs can lead users to copy them verbatim into real deployments, weakening the protection of encrypted configuration.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

This markdown file presents the skill guidance primarily in Chinese, starting with a bilingual heading but immediately switching to Chinese-only instructional content. Under the policy rules, forcing a specific language or locale without user opt-in or a documented region-specific justification is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

Line L24 states that when the user communicates in Chinese, all responses and code comments must be in Chinese. This imposes a language policy automatically rather than offering a user choice, and the file does not document a region-specific or compliance-based justification for the constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

A file that forces a specific language can violate language/locale policy when users are not given an opt-in or alternative. This document consistently uses Chinese for its instructions and examples without stating that it is a region-specific reference or providing language choice.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

This markdown reference is written in Chinese, including its title and usage notes, and does not indicate that users may choose another language. Under the language/locale policy rule, forcing a specific language without user opt-in can be a natural-language policy violation unless the restriction is clearly justified.

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 includes a delete operation example for the distributed file service, but the surrounding documentation does not warn that the action removes stored data and may be irreversible depending on the backend. For markdown files, omissions of warnings about behaviors that can affect user data should be flagged.

Content

No source excerpt is available for this finding.