Official collection of agent skills for TMS Software products and libraries.
Skills are focused instruction packages that AI coding agents load dynamically to perform specialized tasks. By giving your agent the right skill, it gains deep knowledge of TMS APIs, conventions, and patterns — so it writes correct, idiomatic code on the first try instead of hallucinating APIs or guessing at library behavior.
| Skill | Description |
|---|---|
| aurelius-mapping | Map Delphi classes to a relational database using TMS Aurelius ORM attributes |
| aurelius-objects | Save, load, update, delete, merge, and navigate entity objects using TMS Aurelius ORM (TObjectManager) |
Install all TMS skills:
npx skills add tmssoftware/skillsInstall a single skill:
npx skills add https://github.com/tmssoftware/skills/tree/main/skills/aurelius-mappingThe CLI auto-detects your installed agents and places the skills in the right folder. Skills activate automatically when the agent detects a relevant task — no extra configuration needed.
Clone or download this repository, then copy the skill folder(s) into the directory your agent watches:
| Agent | Project-level path | Global path |
|---|---|---|
| Claude Code | .claude/skills/ |
~/.claude/skills/ |
| Cursor | .agents/skills/ |
~/.cursor/skills/ |
| Cline | .cline/skills/ |
~/.cline/skills/ |
| OpenCode | .agents/skills/ |
~/.config/opencode/skills/ |
| Continue | .continue/skills/ |
~/.continue/skills/ |
Example for Claude Code:
# from the repo root
cp -r skills/aurelius-mapping /path/to/your-project/.claude/skills/Each skill is a folder containing:
SKILL.md— trigger conditions, decision rules, and guidance that goes beyond what the reference docs coverreferences/— detailed reference material (attribute syntax, option tables, code examples) that the agent reads on demand
The agent loads the skill when it recognises a matching request, reads the relevant reference material, and applies the rules as it generates code.
TMS libraries like Aurelius, XData, FNC, and others have rich, opinionated APIs. Without accurate context, even capable models make predictable mistakes: wrong attribute names, incorrect object lifetime management, missing registration calls. These skills encode the knowledge needed to avoid those mistakes, making AI-assisted development with TMS products reliable and productive.
Skills follow the Agent Skills specification. To contribute a skill for another TMS product, open a pull request with a new folder under skills/.