No description
- Makefile 100%
- Document lab user create + per-user mycelium bridge - Cover access, stack bring-up from a branch, teardown - Capture toolchain 1.96 and screen/tty start gotchas - List skill in README index |
||
|---|---|---|
| git-work | ||
| hero-lab | ||
| hero-user-provisioning | ||
| mahmoud-hero-devbox | ||
| Makefile | ||
| README.md | ||
claude-skills
Personal Claude Code skills for working on the
Hero ecosystem / mahmoud-ashraf-devbox, plus a standard git workflow. They give
a fresh Claude session the context up front — no need to re-explain how things
work each time.
Skills
| Skill | What it's for |
|---|---|
mahmoud-hero-devbox |
Devbox environment, Hero architecture, lab ops, ecosystem state & known issues |
hero-lab |
Full lab CLI reference (build / service / repo / secrets / install) |
hero-user-provisioning |
Create/manage a multi-tenant Hero user + its own mycelium bridge via lab user / lab flow user |
git-work |
Pre-commit gate (lint → test → build) → branch development_<name> → conventional commit (no Co-Authored-By) → push |
Install
make skills
Symlinks each skill into ~/.claude/skills/<name> so Claude loads them in every
session. Edits in this repo are live (symlinks). Re-run after adding or renaming
a skill. make list shows the skills it will install.
Adding a skill
Create <skill-name>/SKILL.md with YAML frontmatter (name, description),
then make skills. A precise description is what makes Claude auto-load the
skill at the right time.