teaching-skills — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited teaching-skills (Plugin) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Codex-native packaging of the Teaching Skills suite — the full university teaching lifecycle (design → build → assess → deliver → reflect → improve) for OpenAI Codex CLI. This is the sibling Codex distribution of Teaching Skills for Claude Code.
This repository vendors the suite as a single Codex skill that routes to 15 workflows by intent:
skills/teaching-suite/
SKILL.md # router: intent -> workflow, plus ts-* alias emulation
manifest.json # adapter metadata + upstream commit pin
VERSION
agents/openai.yaml
ts/ # vendored content (each upstream SKILL.md -> WORKFLOW.md)
<15 skills>/WORKFLOW.md + agents/ references/ templates/
shared/ scripts/ commands/ MODE_REGISTRY.mdGenerated, not hand-edited. The canonical source is the Claude Code repo; this package is reproduced by scripts/build_codex.py there, with no drift. Do not edit files here directly — change the upstream suite and regenerate.YujxZJCN/teaching-skills — native skills, /plugin install, /ts-* slash commands.
ts-* aliases.Install the skill from this repo with the Codex skill installer:
python3 "$HOME/.codex/skills/.system/skill-installer/scripts/install-skill-from-github.py" \
--repo YujxZJCN/teaching-skills-codex \
--ref main \
--path skills/teaching-suite \
--method gitTo update, remove the installed copy and reinstall:
rm -rf "$HOME/.codex/skills/teaching-suite"
python3 "$HOME/.codex/skills/.system/skill-installer/scripts/install-skill-from-github.py" \
--repo YujxZJCN/teaching-skills-codex \
--ref main \
--path skills/teaching-suite \
--method gitOn macOS / most Linux, Python 3 is python3. If your only python is Python 3, substitute it.
Ask in natural language ("design a course on machine learning for 90 students"), or use a Claude-style alias as a mode shortcut: ts-course, ts-exam, ts-rubric, ts-feedback, ts-cohort, ts-dashboard, and more. The router (skills/teaching-suite/SKILL.md) dispatches to the right workflow and enforces the suite's non-negotiables (professor checkpoints, no invented context, evidence-bound person-affecting output, integrity-by-design, no student data in the passport).
How this works in Codex. Codex does not register Claude Code slash commands. The ts-* names are emulated aliases: the vendored ts/commands/*.md files are prompt recipes the router reads to pick a workflow and mode — they are not installed commands. Type ts-exam ... (with or without a leading slash) and the router treats it as a mode shortcut. If anything looks wrong after an upstream change, recover by reinstalling (the update step above) — the skill is a regenerated build product, so a clean reinstall always restores a consistent state.
The deterministic validators travel too:
python3 skills/teaching-suite/ts/scripts/check_passport.py course_passport.yaml
python3 skills/teaching-suite/ts/scripts/build_dashboard.py course_passport.yamlCodex package 0.2.0. Vendored from YujxZJCN/teaching-skills@50258caf73fb on 2026-06-15. The package version tracks the adapter independently of the upstream suite version; see manifest.json.
MIT — same as the upstream suite. Architecture inspired by academic-research-skills and its Codex sibling.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.