Security audit
Security checks for vulnerabilities and agentic risk
This skill guides focused WelineFramework unit-test work and does not include self-running code or request sensitive access.
Install this if you want an agent to help add focused WelineFramework unit tests. Be aware that it uses Chinese role names and references Weline-specific local guidance files, so it fits best in that project context.
The file uses a Chinese-only skill name and later requires following and notifying Chinese-named roles, but it does not state that the skill is intentionally limited to a Chinese-language environment or offer an alternative language option. That can amount to an implicit locale constraint without user opt-in.
The activation guidance is broad enough that ordinary requests mentioning PHPUnit, Pest, helper tests, or coverage could trigger this specialist skill without a clear boundary check. In a multi-skill agent system, ambiguous routing can cause the wrong skill to take over, pull in extra source material, or apply unintended workflow constraints, increasing the chance of incorrect actions or prompt-scope confusion.
Listing internal skill paths exposes the existence and organization of other agent capabilities. In an adversarial prompt-injection setting, this information can help an attacker map the agent environment, craft more targeted requests, or manipulate routing and cross-skill behavior.
- `AI-ENTRY.md`
- `CLAUDE.md`
- `dev/ai/skills/testing/SKILL.md`
- `dev/ai/skills/service-development/SKILL.md`
- `dev/ai/skills/code-generation-standards/SKILL.md`
- `dev/ai/skills/community-module/SKILLS-CONSOLIDATED.md`
The explicit reference to another internal skill reveals part of the agent's capability graph and filesystem layout. That context can aid reconnaissance by helping an attacker infer available roles, likely prompts, and handoff mechanisms to influence the agent system more effectively.
- `AI-ENTRY.md`
- `CLAUDE.md`
- `dev/ai/skills/testing/SKILL.md`
- `dev/ai/skills/service-development/SKILL.md`
- `dev/ai/skills/code-generation-standards/SKILL.md`
- `dev/ai/skills/community-module/SKILLS-CONSOLIDATED.md`
Exposing the code-generation standards skill path gives attackers another concrete clue about internal controls and where to target prompt manipulation attempts. Even without direct file access, enumerating internal assets lowers uncertainty and improves attack planning against multi-skill orchestration.
- `CLAUDE.md`
- `dev/ai/skills/testing/SKILL.md`
- `dev/ai/skills/service-development/SKILL.md`
- `dev/ai/skills/code-generation-standards/SKILL.md`
- `dev/ai/skills/community-module/SKILLS-CONSOLIDATED.md`
# Responsibilities
The collaboration contract requires adherence to and escalation through Chinese-language role identifiers such as 通用工程师-开发规范与代码质量 and @Weline-技术主管. Without stating that the environment is intentionally Chinese-language or offering equivalent alternatives, this creates a language policy concern.
Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.
- Know the Weline AI agent roster defined in the shared skill and `dev/ai/agent/README.md`.
- Keep work inside this specialist's ownership boundary.
- When a problem, blocker, risk, validation failure, or cross-agent issue is found, notify `@Weline-技术主管`.
- Do not silently expand scope to fix another agent's area.
- Include collaboration status in the final report.
No suspicious patterns detected.