TypeSafe Agent Skills
A skill for building with TypeSafe: typed questions to System One models with calibrated probabilities, not prompt parsing.
- 2,895 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The skill only adds instructions, but code built with it sends data to the external TypeSafe service and needs an API key.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/typesafe-agent-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill only adds instructions, but code built with it sends data to the external TypeSafe service and needs an API key.
This is third-party code. Review the repository files before installing.
What it does
The `typesafe-ai` skill helps design and review code that calls TypeSafe System One models, in particular Jev. The model answers typed questions about text, images and application data: a choice among options, the probability that a statement is true, a score on levels or a box in an image. The agent starts from the decision logic: judgments go to the model while the choice of action stays in code as a truth table. The skill points to the live docs at `docs.typesafe.ai` for exact syntax, SDKs and limits.
Who it is for. Developers replacing a prompt-and-parse step, keyword matching or a brittle heuristic in an application with a typed question to a model.
Good fit when
- Support tickets must be sorted by department and uncertain decisions handed to a human.
- A model reply is parsed with regular expressions and the format breaks from time to time.
- The tone of a message must be scored on a scale or an object found in an image.
- Code already calls TypeSafe and the questions with criteria need checking against live requests.
Not a fit when
- You need ordinary text generation or code rather than classification and probabilities.
- The decision comes down to arithmetic, dates or exact lookup, which is a job for code.
- There is no TypeSafe account or API key and the data cannot go to an external service.
Example request
Use TypeSafe to route incoming support tickets by department, with human review for uncertain decisions.Limitations
The skill only guides the agent: the service itself is paid under TypeSafe terms, an account and API key are needed, and requests with your data go to `api.typesafe.ai`. Exact syntax, limits and model behavior are not baked into the skill and must be read in the live docs, so without network access the agent's advice will be incomplete. Jev is described as the first public System One model and its known weak spots are listed on a separate docs page. The skill is MIT licensed while the service has its own terms.
How to disable. Remove the plugin with `claude plugin uninstall typesafe@typesafe-ai` or the skill through `npx skills remove`, and delete the TypeSafe API key from the environment.
FAQ
How do I call the skill in Claude Code?
With `/typesafe:typesafe-ai` or an ordinary request such as sorting support tickets.
How do I install it outside Claude Code?
Run `npx skills add typesafe-ai/skills --skill typesafe-ai` and pick the agent in the prompt. Installation is project-local and `-g` makes it global.
What are System One models?
TypeSafe models that answer typed questions about application state: a choice, a probability, a score or a box in an image.
Why does the decision stay in code?
The model supplies judgments and code maps them to an action, so the logic is visible and testable.
Is the skill free?
The skill is MIT licensed but the TypeSafe service has its own terms and needs an API key.
Discussion
No questions or reviews yet.
Related
SkillsAn unofficial skill teaching agents to design Jev calls: typed questions, calibrated confidence and examples with code sketches
Official Mastra skills: an agent finds framework APIs in the docs, fixes common errors and walks through version migrations.
A library of 30-plus instruction skills for coding agents: planning, building, testing, hallucination guards and shipping
skill-kit builds agent skills as typed TypeScript state machines with schema-checked answers and a build into a SKILL.md directory
Tech Leads Club Agent Skills
Agent Skills
A vetted skill registry with a CLI that installs into twenty-plus agents and an MCP server for on-demand skill lookup
Agent Skills (AsyrafHussin)
Agent Skills
24 skills for Laravel, PHP, React and TypeScript: code rules, testing, security, technical debt and documentation.
README in short
The repository holds one `typesafe-ai` skill for designing System One integrations. It installs as a Claude Code plugin or through `npx skills add`.