Security audit
Security checks for vulnerabilities and agentic risk
This is a disclosed, prompt-only Chinese fitness coaching skill with safety boundaries and no hidden code, network access, credential handling, or persistence.
Install this only if you want a Chinese-language fitness coaching prompt. Treat its training and nutrition plans as general guidance, not medical care, and avoid sharing sensitive health data unless you are comfortable with the assistant using it in the conversation.
The trigger conditions include very broad phrases such as asking '怎么练/练什么/怎么安排', which can plausibly appear in non-fitness conversations and cause unintended activation. In a skill that provides health, exercise, and diet guidance, misrouting a user into the wrong workflow can produce irrelevant or unsafe advice if the surrounding context is not actually about fitness.
The trigger scope for diet terms and body signals is broad enough that ordinary discussion of food, pain, or medical symptoms could invoke this coaching skill without sufficient domain confirmation. Because this skill can proceed into screening and training/nutrition guidance, an accidental match in a borderline medical context could lead to inappropriate quasi-medical responses instead of deferring safely.
L119 明确规定“中文”，属于对输出语言的强制要求。文件中未说明这是面向特定中文用户群的限定，也未给出用户可选择其他语言的选项，因此符合语言/locale 策略违规特征。
Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.
cp -r koo-fitness-coach ~/.workbuddy/skills/
# Claude Code（项目级）
mkdir -p .claude/skills && cp -r koo-fitness-coach .claude/skills/
# Hermes Agent
cp -r koo-fitness-coach ~/.hermes/skills/
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.
cp -r koo-fitness-coach ~/.workbuddy/skills/
# Claude Code（项目级）
mkdir -p .claude/skills && cp -r koo-fitness-coach .claude/skills/
# Hermes Agent
cp -r koo-fitness-coach ~/.hermes/skills/
This skill's main description, workflow, routing, output contract, and operating rules are primarily presented in Chinese, which can force a specific language experience on users who did not opt into that locale. Although brief English fields exist, the operative instructions remain Chinese-only, creating a language/locale policy concern.
This markdown file contains user-facing assessment instructions exclusively in Chinese, and there is no indication that the user can choose another language or that the skill is intentionally limited to a Chinese-speaking or region-specific audience. That can violate language/locale policy when users are not given an explicit opt-in to a specific language.
This file gives specific calorie-deficit targets, weight-loss rates, and exercise prescriptions, including guidance for postpartum and other higher-risk users, but it does not clearly state that the content is general information and not individualized medical advice. In a fitness-coach skill, users may reasonably treat these recommendations as personalized and safe, which creates risk of inappropriate dieting or training for people with medical conditions, injury history, eating disorders, or inadequate recovery capacity.
The file explicitly claims it only serves healthy users and that chronic disease, eating disorder history, pregnancy, lactation, and other special cases must be referred out, but later sections still give concrete advice for groups such as hypertension, diabetes, and kidney disease. This contradiction can cause the agent to overstep its stated safety boundary and provide quasi-clinical nutrition or hydration guidance to users who should instead receive individualized medical advice.
The 'always refer' statement is undermined by later actionable special-population guidance, creating ambiguous behavior for the agent. In a fitness-coach skill, that ambiguity is dangerous because users may reasonably rely on the advice as personalized health guidance despite the initial disclaimer, especially around hydration, sodium, glucose, and renal limits.
The entire skill reference is written in Chinese and does not indicate that language selection is optional or user-configurable. Under the policy criteria, forcing a specific language without user opt-in is a natural-language locale violation unless the constraint is documented and justified.
The file title and all instructional content are written in Chinese, and line L54 states the weekly plan output format 'must' follow the provided structure, which is also entirely in Chinese. There is no indication that the user may choose another language or that the locale restriction is justified, which can violate language/locale policy requirements.
This markdown file presents all instructions and examples exclusively in Chinese, with no user opt-in, alternative language option, or justification for a Chinese-only audience. Under the policy rule for natural-language violations, forcing a single language without documented choice can be a locale-policy issue.
This markdown file presents all guidance in Chinese and does not offer an alternative language or indicate that the user has opted into that locale. Under the policy, forcing a specific language without user opt-in is a natural-language policy concern.
This markdown skill content forces a specific language for all users, which can violate language or locale policy when no opt-in or alternative is provided. There is no indication that the skill is region-specific or that users can select another language.
This file contains user-facing natural-language instructions exclusively in Chinese, which can amount to a language policy issue if users are not given an opt-in or alternative locale. The document does not state that the skill is intended only for Chinese-speaking users or provide any language selection.
This file presents all operational safety guidance exclusively in Chinese, and there is no indication that users can choose another language or that the skill is intentionally limited to a Chinese-speaking region. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy concern.
This markdown file is entirely presented in Chinese and does not indicate that the user can choose another language or that the skill is intentionally limited to Chinese-speaking users. Under the policy criteria, forcing a specific language without user opt-in is a natural-language policy violation.