扫描报告
5 /100
lucid-skill
AI-native data analysis skill. Connect Excel/CSV/MySQL/PostgreSQL, understand business semantics, query with natural language.
Legitimate AI-native data analysis skill with proper security controls including SQL safety whitelisting, credential stripping, and read-only query enforcement.
可以安装
This skill is safe to use. No security concerns identified.
安全发现 1 项
| 严重性 | 安全发现 | 位置 |
|---|---|---|
| 低危 | Loose dependency version constraints 供应链 | pyproject.toml:18 |
| 资源类型 | 声明权限 | 推断权限 | 状态 | 证据 |
|---|---|---|---|---|
| 文件系统 | READ | READ | ✓ 一致 | User-provided CSV/Excel files only, stored in ~/.lucid-skill/ |
| 网络访问 | READ | READ | ✓ 一致 | Downloads ~460MB embedding model from HuggingFace (declared in SKILL.md) |
| 命令执行 | NONE | NONE | — | No subprocess or os.system calls found |
| 环境变量 | READ | READ | ✓ 一致 | Reads LUCID_DATA_DIR and LUCID_EMBEDDING_ENABLED only |
| 技能调用 | ADMIN | ADMIN | ✓ 一致 | Full MCP tool access (connect, query, search, etc.) |
| 数据库 | READ | READ | ✓ 一致 | MySQL/PostgreSQL connections for schema discovery and SELECT queries only |
目录结构
63 文件 · 227.2 KB · 6647 行 Python 45f · 5544L
Markdown 7f · 791L
CSV 9f · 191L
YAML 1f · 89L
TOML 1f · 32L
├─
▾
lucid_skill
│ ├─
▾
catalog
│ │ ├─
__init__.py
Python
│ │ ├─
profiler.py
Python
│ │ ├─
schema.py
Python
│ │ └─
store.py
Python
│ ├─
▾
connectors
│ │ ├─
__init__.py
Python
│ │ ├─
base.py
Python
│ │ ├─
csv_conn.py
Python
│ │ ├─
excel_conn.py
Python
│ │ ├─
mysql_conn.py
Python
│ │ └─
postgres_conn.py
Python
│ ├─
▾
discovery
│ │ ├─
__init__.py
Python
│ │ ├─
domains.py
Python
│ │ └─
joins.py
Python
│ ├─
▾
query
│ │ ├─
__init__.py
Python
│ │ ├─
engine.py
Python
│ │ ├─
formatter.py
Python
│ │ ├─
router.py
Python
│ │ └─
safety.py
Python
│ ├─
▾
semantic
│ │ ├─
__init__.py
Python
│ │ ├─
embedder.py
Python
│ │ ├─
hybrid.py
Python
│ │ ├─
index.py
Python
│ │ ├─
layer.py
Python
│ │ └─
search.py
Python
│ ├─
▾
tools
│ │ ├─
__init__.py
Python
│ │ ├─
connect.py
Python
│ │ ├─
describe.py
Python
│ │ ├─
discovery.py
Python
│ │ ├─
overview.py
Python
│ │ ├─
profile.py
Python
│ │ ├─
query.py
Python
│ │ ├─
search.py
Python
│ │ └─
semantic.py
Python
│ ├─
__init__.py
Python
│ ├─
__main__.py
Python
│ ├─
cli.py
Python
│ ├─
config.py
Python
│ ├─
server.py
Python
│ ├─
startup.py
Python
│ └─
types.py
Python
├─
▾
references
│ ├─
commands.md
Markdown
│ ├─
json-schema.md
Markdown
│ └─
workflow.md
Markdown
├─
▾
semantic_store
│ └─
▾
csv_orders_csv
│ └─
orders.yaml
YAML
├─
▾
tests
│ ├─
▾
datasets
│ │ ├─
▾
cross-source-a
│ │ │ └─
shop_orders.csv
CSV
│ │ ├─
▾
cross-source-b
│ │ │ └─
shop_customers.csv
CSV
│ │ ├─
▾
ecommerce
│ │ │ ├─
customers.csv
CSV
│ │ │ ├─
order_items.csv
CSV
│ │ │ ├─
orders.csv
CSV
│ │ │ └─
products.csv
CSV
│ │ ├─
▾
hr
│ │ │ ├─
departments.csv
CSV
│ │ │ └─
employees.csv
CSV
│ │ ├─
▾
superstore
│ │ │ └─
orders.csv
CSV
│ │ └─
README.md
Markdown
│ ├─
test_catalog.py
Python
│ ├─
test_config.py
Python
│ ├─
test_csv_connect.py
Python
│ ├─
test_engine.py
Python
│ └─
test_safety.py
Python
├─
CLAUDE.md
Markdown
├─
pyproject.toml
TOML
├─
README.md
Markdown
└─
SKILL.md
Markdown
依赖分析 7 项
| 包名 | 版本 | 来源 | 已知漏洞 | 备注 |
|---|---|---|---|---|
duckdb | >=1.0 | pip | 否 | Minimum version specified |
click | >=8.0 | pip | 否 | Minimum version specified |
pyyaml | >=6.0 | pip | 否 | Minimum version specified |
mcp | >=1.0 | pip | 否 | Minimum version specified |
psycopg2-binary | unpinned | pip[db] | 否 | Optional dependency |
mysql-connector-python | unpinned | pip[db] | 否 | Optional dependency |
sentence-transformers | unpinned | pip[embedding] | 否 | Optional dependency |
安全亮点
✓ SQL safety checker uses whitelist approach - only SELECT/WITH allowed
✓ Credentials explicitly stripped before persistence (store.py:upsert_source)
✓ No shell execution or subprocess calls found
✓ No credential harvesting or exfiltration to external IPs
✓ No obfuscation or base64-encoded payloads
✓ Read-only queries enforced - INSERT/UPDATE/DELETE/DROP blocked
✓ Comprehensive test coverage for safety checker
✓ MIT licensed open source with transparent implementation
✓ DuckDB used for both catalog storage and query engine (sandboxed)