agent-skills-216ddf — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited agent-skills-216ddf (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.
Reusable agent skills for Codex, Claude Code, OpenCode, and other agents that support the open agent skills format.
The canonical skill source is the top-level skills/ directory. Plugin manifests and marketplace files point at the same source so the repository can be installed through multiple agent ecosystems.
docs-sync: Sync documentation with user-facing code changes, or check whether docs define intended code behavior.flutter-flavors: Set up audit Flutter flavors, flutter_flavorizr / flavorizr.yaml, platform app identities, launch configs, and build-mode boundaries.git-workflow: Guide safe local Git workflows such as staging, commits, branches, merges, rebases, tags, and recovery.github-workflow: Guide GitHub collaboration workflows such as pull requests, review threads, Actions checks, releases, and gh CLI usage.github-upstream-sync: Create or review GitHub Actions workflows that sync fork repositories from upstream.$docs-sync Check whether the docs need updates for this diff.$flutter-flavors Audit Android/iOS flavors and check whether flavorizr.yaml matches native files.$git-workflow Help me split these changes into safe commits.$github-workflow Review this PR's checks and merge readiness.$github-upstream-sync Create a workflow that syncs my repository's main branch from upstream main every day at 4:00.Run this from the project folder where you want the skills installed:
npx skills@latest add mabyko/AgentSkillsBy default, npx skills add installs per project. Use --global only when you want a user-level install.
For projects that already installed these skills with the skills CLI, use the update command:
npx skills@latest updateskills update reads the installed skill lock files, fetches the latest source, removes the old installed skill, and reinstalls the updated version. This keeps installed skills aligned with upstream changes, including files removed from references/ or other bundled resource folders.
Use add for first-time installation:
npx skills@latest add mabyko/AgentSkillsList available skills without installing:
npx skills@latest add mabyko/AgentSkills --listInstall a specific skill:
npx skills@latest add mabyko/AgentSkills --skill docs-syncInstall for specific agents:
npx skills@latest add mabyko/AgentSkills -a claude-code -a codex -a opencodeInstall globally:
npx skills@latest add mabyko/AgentSkills --globalThe skills CLI discovers this repository's top-level skills/ directory and installs each selected skill into the target agent's expected location.
Use this path when you want Codex to install the repository as a plugin from a marketplace.
codex plugin marketplace add mabyko/AgentSkillsThen open Codex, go to /plugins, search for agent-skills, and install it.
Codex resolves this repository through:
.agents/plugins/marketplace.json
└── source.path: "./"
└── .codex-plugin/plugin.json
└── skills: "./skills/"The marketplace file is the catalog entry. .codex-plugin/plugin.json is the plugin manifest. Skills remain in skills/.
Use this path when you want Claude Code to install the repository as a plugin.
/plugin marketplace add mabyko/AgentSkills
/plugin install agent-skills@agent-skillsClaude Code reads .claude-plugin/plugin.json as the plugin manifest and uses the repository's top-level skills/ directory as the skill source.
Note: Plugin installs may be cached by the host tool. If you need the latest skills, refresh, update, or reinstall the plugin through that tool's plugin manager.
CLAUDE.md contains:
@AGENTS.mdThis keeps Claude Code's repository guidance aligned with canonical instructions in AGENTS.md.
skills/
└── <skill-name>/
├── SKILL.md
├── agents/openai.yaml
├── references/
├── scripts/
└── assets/
.codex-plugin/
└── plugin.json
.claude-plugin/
└── plugin.json
.agents/
└── plugins/marketplace.json
templates/
└── skill/
scripts/
├── new-skill.sh
└── validate-skills.sh
AGENTS.md
CLAUDE.mdCreate every new skill under the top-level skills/ directory. Do not add new skills under .agents/skills/, .claude/skills/, or plugins/.
Start a new skill with:
scripts/new-skill.sh my-skillThis creates:
skills/my-skill/
├── SKILL.md
└── agents/openai.yamlBefore opening a pull request:
SKILL.md with a clear name, trigger-focused description, and concise workflow steps.agents/openai.yaml with a display name, short description, brand color, and default_prompt that mentions $my-skill.references/.scripts/ directory.assets/ directory.scripts/validate-skills.shSkill names must use kebab-case, such as release-notes, frontend-review, or python-debugging.
Each skill is a folder containing:
SKILL.md: requiredagents/openai.yaml: optional OpenAI/Codex UI metadata and dependenciesscripts/: optional deterministic helpersreferences/: optional docs loaded only when neededassets/: optional templates, images, or other files used by the skillMIT
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.