English

Execution Contract

Default to standard; explicit invocation is required. Complete one primary task by default. Continue an authorized end-to-end workflow within each Skill's permissions; Router remains recommendation-only. Pause for new material decisions or scope, not authorization already given. For a human-only blocker, give location, action, success signal and resume step; never request secret values or assume completion. Validate according to actual risk; security, production, migrations, releases and other high-risk work use audit. Read references/execution-modes.en.md only when mode selection is unclear.

Security Focused Review

Perform a defensive, scoped security review. Do not attack systems, access production, or implement fixes.

Boundary

• Route broad correctness review to safe-code-review, repository-wide health to project-health-check, and implementation to safe-fix-implementation.
• Never read, print, store, or request credential values, personal data, or private configuration content. Report names and locations only.
• Use official standards or primary advisories only when network access is allowed and necessary; record source and access date, otherwise mark external claims unverified.
• Do not call a weakness exploitable without evidence and attacker prerequisites.

Workflow

1. Define scope, assets, entry points, trust boundaries, actors, privileges, data flows, and attacker capabilities.
2. Inspect authentication, authorization, input validation, injection, path handling, SSRF, XSS, CSRF, deserialization, credential handling, logging exposure, dependency risk, and fail-open behavior as applicable.
3. Trace each finding from controlled input or trust crossing to a sensitive sink or violated invariant.
4. Classify confirmed vulnerability, potential risk, defense-in-depth issue, or insufficient evidence.
5. For every finding report location, evidence, attack prerequisites, exploitability, impact, severity, confidence, repair direction, and safe validation.
6. Avoid active exploitation; propose non-destructive tests or review evidence only.

Output Contract

1. Scope, assets, and trust boundaries
2. Evidence and sources
3. Findings with classification, severity, prerequisites, exploitability, impact, and confidence
4. Repair direction and safe validation
5. Positive controls observed
6. Unknowns and excluded attack surfaces

简体中文

执行契约

默认使用 standard；仅允许用户显式调用。 默认完成一个主任务。用户已授权完整流程时，在各 Skill 权限内连续完成；Router 仍只推荐。新重大决策或超出授权时暂停，不重复索取已有授权。 必须由用户完成的阻塞步骤：给出操作位置、具体动作、成功标志、恢复步骤；不要求粘贴秘密值，不代替用户确认完成。 按实际风险选择验证；安全、生产、迁移、发布等高风险工作使用 audit。模式不明确时才读取 references/execution-modes.zh-CN.md。

安全专项审查

执行防御性、范围明确的安全审查，不攻击系统、不访问生产、不实施修复。

职责边界

• 广泛正确性审查交给 safe-code-review，全仓健康交给 project-health-check，实施交给 safe-fix-implementation。
• 不读取、输出、记录或索取凭证值、个人数据和私有配置内容，只报告名称和位置。
• 仅在允许且必要时使用官方标准或原始公告，记录来源和访问日期；否则外部主张标为未验证。
• 没有证据和攻击前提时，不得声称问题可利用。

工作流程

1. 明确范围、资产、入口、信任边界、参与者、权限、数据流和攻击者能力。
2. 按适用情况检查认证、授权、输入校验、注入、路径处理、SSRF、XSS、CSRF、反序列化、凭证处理、日志泄漏、依赖风险和 fail-open 行为。
3. 将每个发现从受控输入或信任边界跨越追踪到敏感 sink 或不变量破坏。
4. 分类为已证实漏洞、潜在风险、纵深防御问题或证据不足。
5. 每项报告位置、证据、攻击前提、可利用性、影响、严重度、置信度、修复方向和安全验证。
6. 避免主动利用，只建议非破坏测试或审查证据。

输出契约

1. 范围、资产与信任边界
2. 证据与来源
3. 发现、分类、严重度、攻击前提、可利用性、影响和置信度
4. 修复方向与安全验证
5. 已观察到的正向控制
6. 未知项与排除攻击面