# Orlix Session Lessons

> Use when deriving durable Orlix lessons from a conversation, correcting agent behavior, or improving harness reliability.

- Skill: `rudironsoni/orlix-session-lessons` (Agent Skill)
- Install (CLI): `npx skillmds@latest add rudironsoni/orlix-session-lessons`
- Raw SKILL.md: https://api.skillmd.com/api/skills/rudironsoni/orlix-session-lessons/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: rudironsoni (https://skillmd.com/u/rudironsoni)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/rudironsoni/orlix-session-lessons

---

# Orlix Session Lessons

Read `AGENTS.md`, `docs/index.md`, `docs/ontology.md`, the relevant canonical pages, current structured reports, and recent commits before proposing a durable lesson.

- Put stable domain knowledge in the owning object or concept page.
- Put a repeated workflow correction in the narrowest skill, hook, test, or RuleSync source that can enforce it.
- Record the knowledge operation in `docs/log.md` and regenerate `docs/index.md`.
- Keep current status, command transcripts, failures, and remaining tasks out of documentation.
- Do not turn one conversation into broad policy without repository evidence.
- Do not duplicate canonical architecture facts across skills and object pages.