FitDietKernel

健身饮食的个人知识库 - 让AI精准帮你算饮食（碳循环/生酮/增肌/减脂）

Security Scan

The skill's files, instructions, and requirements are coherent with a local personal diet knowledge-base: it reads/writes local JSON/MD files and provides desktop UIs and report generation, and it does not request unrelated credentials or network installs.

Purpose & Capability

Name/description match the included files: knowledge_base JSON/MD, profile.json, GUIs, and a report generator. No unrelated credentials, binaries, or platform APIs are required. The code and docs are all about diet tracking and macros.

Instruction Scope

SKILL.md explicitly tells the agent to read the local repository (profile.json, knowledge_base/*.json, logs/*.md) and to update/register new foods. Reading and (per instructions) updating local files is expected for a local knowledge-base, but it means the agent needs filesystem access and is expected to write changes to that repo. Also: SKILL.md/logic_rules state the AI will 'auto update food_registry.json' but the shipped Python scripts do not include an explicit API to perform that edit — the update would be performed by whatever agent code you allow to write files.

Install Mechanism

There is no install spec in the registry (instruction-only), but the bundle includes code and a README that suggests git clone or an npx clawhub installer. No remote download/install URLs or extract-from-URL steps are present in the package itself; risk from install mechanism is low if you clone from a trusted source. If you use the npx method, verify the clawhub package and the source repo first.

Credentials

The skill requests no environment variables, credentials, or config paths. Its data access is limited to local files in the repo (profile, knowledge_base, logs, output reports), which is proportionate to a personal diet knowledge-base.

Persistence & Privilege

The skill is not marked 'always'. It expects to read and (per docs) write local files (food_registry.json, logs, weekly_report). This is a normal level of privilege for a local knowledge-base, but installing it will put code and data on disk and the AI (if granted filesystem write) could modify those files.

Assessment

This skill appears to be what it claims: a local personal diet knowledge base with JSON/markdown data, small GUI scripts, and a report generator. Before installing: 1) verify the repository source (README references a GitHub repo); 2) be aware the skill expects the agent to read and possibly write files in the repository (profile.json, food_registry.json, logs, weekly_report) — if you don't want automatic edits, don't grant write access to the agent or back up the folder; 3) the SKILL.md/logic files say the AI will 'auto-update' the food registry, but that behavior is an instruction-level expectation (no dedicated update API is bundled), so review any agent code that will perform those edits; 4) there are no hidden network endpoints or credential requests in the provided files, but if you install via third-party installers (npx/clawhub) confirm those packages and sources first; 5) running GUI/desktop scripts should be done in a safe environment if you are unsure (e.g., isolated VM) — otherwise the skill is coherent and low risk.

健身饮食的个人知识库，让 AI 成为你的精准饮食教练

功能

• 🔢 精准碳水计算（按体重动态计算）
• 🍚 食物注册（10g 碳水等值换算）
• 📊 误差追踪（±10g 安全线）
• 📋 自动打卡记录
• 📈 周报生成

计算公式

碳循环

低碳日 = 体重 × 1.0g
中碳日 = 体重 × 1.5g
高碳日 = 体重 × 3.0g

蛋白 = 体重 × 1.5~2.0g
脂肪 = 剩余热量

食物换算 (10g 碳水)

食材	分量
燕麦	16g (干)
意面	13g (干)
杂粮馒头	0.5个
蓝莓	72g
红薯	50g

使用方法

1. 克隆仓库到本地
2. 修改 profile.json 配置你的体重、目标
3. 告诉 AI 知识库路径，让它读取数据

你是一个精准饮食教练，接入了 FitDietKernel。
数据路径：~/fitdietkernel/