Schema

Design and manage database schemas using CLI tools. Use when you need to create tables, columns, indexes, relations, validate, migrate, diff,

Security Scan

The skill's description promises a full schema design/migration tool, but the bundled script largely only logs commands to ~/.schema/*.log and several documentation/prerequisite claims don't match the actual code.

Purpose & Capability

SKILL.md describes a feature-rich schema tool (data persisted as ~/.schema/data.jsonl, migration generation, validation, import/export, dialect-aware SQL), but the included script only appends timestamped entries to multiple .log files (create.log, table.log, etc.) and does not implement real migration generation, validation, or JSONL storage. The SKILL.md lists Python 3.8+ as a prerequisite, but the script is pure Bash and does not require Python. These mismatches indicate the implementation is incomplete or misleading relative to the stated purpose.

Instruction Scope

Runtime instructions tell the agent to run scripts/script.sh for all operations, which is consistent; however the SKILL.md instructs imports/exports and dialect-specific SQL output while the script's 'import' handler merely logs the import command and does not parse files, and the intended cmd_export function is present but appears unreachable due to duplicate 'export' case labels (first 'export' case only logs the action). The instructions therefore can mislead the agent into thinking operations are performed when they only get recorded locally.

Install Mechanism

No install spec; this is instruction-only with a bundled Bash script. No network downloads or external install steps are present. The script writes to the user's home directory (~/.schema) which is expected for a local CLI tool.

Credentials

The skill requests no environment variables or credentials. It uses HOME to place data under ~/.schema — expected for a local CLI tool. There is no network access or external credential use in the code.

Persistence & Privilege

The skill is not always-enabled and does not request elevated privileges. It persists data under the user's home directory (~/.schema) and does not modify other skills or system-wide configuration.

What to consider before installing

This skill is inconsistent: documentation promises JSONL storage, Python, and full migration/validation features, but the script only records commands to ~/.schema/*.log and doesn't perform the advertised operations. Before installing or invoking, review the script yourself and consider: 1) Back up or inspect ~/.schema if it already exists (the tool will create and write files there). 2) Test the script in an isolated environment or container to confirm actual behavior. 3) If you expect real migration/validation/import functionality, ask the author for a corrected implementation (or patches) — the current code appears incomplete/buggy (duplicate 'export' case, import is a no-op, docs out of sync). There is no evidence of network exfiltration or credential access, but the mismatch between claims and code is a red flag; treat the skill as untrusted until the implementation matches its description.

Schema — Database Schema Design Tool

A thorough CLI tool for designing, managing, and evolving database schemas. Supports table creation, column management, indexing, relationship mapping, validation, migration generation, schema diffing, and import/export — all stored locally in JSONL format.

Prerequisites

• Python 3.8+
• Bash shell
• jq (optional, for pretty JSON output)

Data Storage

All schema data is persisted in ~/.schema/data.jsonl. Each line is a JSON object representing a schema entity (table, column, index, or relation). This enables easy versioning, diffing, and portability.

Commands

Run all commands via the script at scripts/script.sh.

create

Create a new schema/database namespace.

bash scripts/script.sh create <schema_name> [--charset utf8mb4] [--engine InnoDB]

table

Create or manage a table within a schema.

bash scripts/script.sh table <schema_name> <table_name> [--if-not-exists]

column

Add, modify, or remove a column from a table.

bash scripts/script.sh column <schema_name> <table_name> <column_name> <type> [--nullable] [--default value] [--primary] [--unique] [--remove]

index

Create or drop an index on a table.

bash scripts/script.sh index <schema_name> <table_name> <index_name> <columns> [--unique] [--drop]

relation

Define a foreign-key relationship between tables.

bash scripts/script.sh relation <schema_name> <from_table.column> <to_table.column> [--type one-to-many] [--on-delete CASCADE]

validate

Validate a schema for integrity issues (missing references, orphan columns, duplicate indexes).

bash scripts/script.sh validate <schema_name>

migrate

Generate SQL migration statements from current schema state.

bash scripts/script.sh migrate <schema_name> [--format sql|json] [--dialect mysql|postgres|sqlite]

diff

Compare two schemas and show differences.

bash scripts/script.sh diff <schema_a> <schema_b>

export

Export a schema to JSON or SQL format.

bash scripts/script.sh export <schema_name> [--format json|sql|yaml]

import

Import a schema definition from a JSON file.

bash scripts/script.sh import <file_path> [--schema_name override_name]

help

Show usage information and available commands.

bash scripts/script.sh help

version

Show the current version of the schema tool.

bash scripts/script.sh version

Workflow Example

# Create a schema
bash scripts/script.sh create myapp --charset utf8mb4

# Add tables
bash scripts/script.sh table myapp users
bash scripts/script.sh table myapp posts

# Add columns
bash scripts/script.sh column myapp users id INTEGER --primary
bash scripts/script.sh column myapp users email VARCHAR --unique
bash scripts/script.sh column myapp posts id INTEGER --primary
bash scripts/script.sh column myapp posts user_id INTEGER

# Add relation
bash scripts/script.sh relation myapp posts.user_id users.id --type one-to-many

# Validate
bash scripts/script.sh validate myapp

# Generate migration
bash scripts/script.sh migrate myapp --dialect postgres

Notes

• All data is local — no database connection required for design phase.
• Schema definitions are portable via export/import.
• Validation catches common design issues before migration.
• Supports MySQL, PostgreSQL, and SQLite dialects for migration output.

---

Powered by BytesAgain | bytesagain.com | [e-mail]