# Agent skill

For Claude Code and other coding agents: a skill file, `SKILL.md`, tells an agent what Jepela is, when to use it inside the software it writes, and the exact request shape. Get it from [https://github.com/ceyoualigator-debug/jepela](https://github.com/ceyoualigator-debug/jepela) (`skills/jepela/SKILL.md`) and copy it into `.claude/skills/jepela/` of a project, or into `~/.claude/skills/jepela/` for every project. The plugin brings a different skill, `jepela-decisions`, for Claude's own tool calls.

The short version an agent needs:

- Jepela is not a chat model and does not write code. It answers typed questions about a state: choice, score, noul, with probabilities and confidence, in one request.
- Use it inside applications for routing, classification, scoring, yes/no checks, and for decisions that should take a subject's memory into account.
- `POST /v1/systemone` with `state`, `questions`, optional `subject`; keys in `JEPELA_API_KEY`; the base URL is `https://api.jepela.com`.
- Ask one thing per question; combine answers in code; gate actions on confidence; remove options that must never be chosen in code (an exclusion rule in memory, or leave them out of the question); pin facts that must count at every decision; put numeric and date conditions in code (`derive`, or a memory rule with `when`).
- Read `/llms.txt` on the gateway for the documentation index.

For Claude Code there is also a plugin with an MCP server, so Claude can call Jepela itself: `/plugin marketplace add ceyoualigator-debug/jepela`, then `/plugin install jepela@jepela`. See Claude and MCP.