WorkBuddy App Builder Skill
A WorkBuddy skill that turns a business description into an ontology and generates a new skill with SQLite, an API and forms.
- 64 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The generated server gives read and write access to the database with no sign-in, and `/api/sql` executes queries against it.
- The engine starts and writes files inside the skill directory.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/workbuddy-app-builder-skillDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The generated server gives read and write access to the database with no sign-in, and `/api/sql` executes queries against it.
- The engine starts and writes files inside the skill directory.
This is third-party code. Review the repository files before installing.
What it does
The `ontology-app-builder` skill takes a business idea or requirements document to a working application. It breaks requirements into nine stages and asks for confirmation at each, then builds an ontology of eleven models, M1–M7, ME, MU, MM and MI, in YAML. The ontology appears once all questions are closed, and the domain skill once the model is confirmed. The resulting domain skill is self-contained. It includes a requirements document, ontology JSON, a Python engine with a SQLite database, a REST API for create, read, update and delete, standalone HTML forms for each aggregate root and an offline ECharts knowledge graph. WorkBuddy turns natural-language questions into a read-only `SELECT` after getting the database structure from `/api/schema`.
Who it is for. Analysts and builders of internal systems on WorkBuddy who need a quick prototype of a record-keeping app with a confirmable domain model.
Good fit when
- You need an application for supplier admission and evaluation built from a process description.
- Contract requirements become a model of objects, rules and roles.
- Data lives in a local SQLite database and input goes through standalone HTML forms.
- A manager asks questions about the data in plain words and the answer comes from the database.
- Relations between domain objects must be shown as an offline knowledge graph.
Not a fit when
- You do not work in WorkBuddy: the skill's tool paths are tied to `~/.workbuddy/skills/`.
- You need a multi-user system with sign-in, roles and access control: the engine has none.
- The application goes onto the internet: the service listens on `127.0.0.1` only and is not built for public access.
Example request
Help me build an application for supplier admission and performance evaluation: ask the requirements questions first.Limitations
The project is a local single-user MVP. The engine has no sign-in, tenant isolation or permission checks, and event and approval models are built but not executed. Entry of main and child objects goes one nesting level deep. The database structure is created with `CREATE TABLE IF NOT EXISTS` and no migrations, so after model changes back up the data first. Building needs Python 3 and PyYAML, while the finished domain skill runs on the standard library. The skill directory must be named `ontology-app-builder`. Documentation is mainly Chinese with an English version.
How to disable. Delete the `~/.workbuddy/skills/ontology-app-builder` directory and reload the WorkBuddy skill list. Generated domain skills and their `data/app.db` databases sit separately and are deleted by hand.
FAQ
How do I install the skill in WorkBuddy?
Clone the repository and copy it to `~/.workbuddy/skills/ontology-app-builder`, for example `cp -R ./WorkBuddy-AppBuilderSkill/. ~/.workbuddy/skills/ontology-app-builder/`. Then reload the skill list and confirm the skill is enabled.
What does a generated domain skill contain?
A requirements document, ontology YAML and JSON, an engine with SQLite, a REST API, HTML forms per root object, plain-language queries and an offline knowledge graph. The Python standard library is enough to run it.
How do I start the generated engine by hand?
In the domain skill root run `python engine/run.py --port 8990`. The server listens on `http://127.0.0.1:8990`, creates `data/app.db` and serves pages and the API from the ontology.
How do plain-language queries work?
The server does not parse them. WorkBuddy gets the database structure from `/api/schema`, writes a read-only `SELECT`, submits it to `/api/sql` and formats the result as a table or an answer.
Can I expose the application to the internet?
No, you should not: the server has no sign-in, listens on `127.0.0.1` only and its SQL check is basic, not a sandbox for outside users.
Discussion
No questions or reviews yet.
Related
SkillsBuilds a business system from a requirements text: discovery, seven ontology YAML models, then an app with manual sign-off.
skill-creator: skills from OpenAPI, GraphQL and MCP
@asnd/skill-creator
The /skill-creator command turns a link to an OpenAPI spec, GraphQL schema or MCP server into a ready skill with wrappers.
A plugin and skill for OpenCode that draft a skill, build eval queries, measure description trigger accuracy and improve it.
A skill that draws a clickable Mermaid diagram in HTML: how a feature, a data flow or database tables work, with code per node
Orval skill for agents
Orval skill
The official Orval skill: the agent sets up generation of TypeScript clients, Query hooks, Zod schemas and MSW mocks from an OpenAPI spec
Builder Skills
Skills for agents
A set of small skills from Builder.io for agent work habits: visual plans and diff recaps, docs-first checks, cross-agent review and clean status updates
README in short
The author contrasts the skill with the usual chain of interviews, modeling, API and forms done one after another. Here the whole chain runs under human control at key steps and the result runs at once.