Skillget

mode-io-skills

agent
Guvenlik Denetimi
Gecti
Health Gecti
  • License — License: Apache-2.0
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Community trust — 12 GitHub stars
Code Gecti
  • Code scan — Scanned 12 files during light audit, no dangerous patterns found
Permissions Gecti
  • Permissions — No dangerous permissions requested

Bu listing icin henuz AI raporu yok.

SUMMARY

Agent skills for Modeio anonymization and safety checks across AI coding assistants.

README.md

Mode IO logo

Mode IO Skills 🛡️🔍

🔒 Privacy, 🛡️ safety, and 🔍 repository-audit skills for AI agents.

Website GitHub Apache 2.0 Visitor badge

Mode IO helps agents do three things well:

  • protect sensitive data before it leaves the prompt
  • check risky instructions before tools or state changes run
  • audit third-party skills and plugins before install

This repository is the home of three active skills:

Skill What it does
privacy-protector Anonymizes and restores PII in text and supported files, with local detector tuning and higher-assurance API-backed modes
security Runs live safety checks on instructions that may trigger tools, edits, destructive actions, or compliance-sensitive operations
skill-audit Performs deterministic static safety audits for third-party skill and plugin repositories before install or execution

Why Teams Use It

  • Safer prompts: redact sensitive data before it reaches shared channels or external models
  • Safer execution: stop risky instructions before they become destructive actions
  • Safer installs: screen third-party skills and plugins with evidence-backed static analysis
  • Multi-agent ready: designed for Claude Code, Codex CLI, OpenCode, and OpenClaw workflows

See It In Action

privacy-protector

cd privacy-protector
python3 scripts/anonymize.py \
  --input "Name: John Doe, Email: [email protected], SSN: 123-45-6789" \
  --level lite

Name: [NAME_1], Email: [EMAIL_1], SSN: [SSN_1]

security

cd security
python3 scripts/safety.py \
  -i "Drop all tables in the production database" \
  -c '{"environment":"production","operation_intent":"destructive","scope":"broad","data_sensitivity":"regulated","rollback":"none","change_control":"ticket:DB-9021"}' \
  -t "postgres://prod/main" \
  --json

{
  "approved": false,
  "risk_level": "critical"
}

skill-audit

cd skill-audit
python3 scripts/skill_safety_assessment.py evaluate \
  --target-repo /path/to/repo \
  --json

decision: caution
risk_score: 42

Install

Install only the skill you need.

Option 1: ClawHub / OpenClaw

When a skill is listed in ClawHub, install the specific one you need:

clawhub install privacy-protector
clawhub install security
clawhub install skill-audit

Option 2: npx skills add

If you prefer the repo-path workflow, npx skills add is still supported:

npx skills add mode-io/mode-io-skills --skill privacy-protector --agent codex --yes --copy
npx skills add mode-io/mode-io-skills --skill security --agent codex --yes --copy
npx skills add mode-io/mode-io-skills --skill skill-audit --agent codex --yes --copy

Swap codex for claude-code or opencode as needed.

Learn More

License

Apache License 2.0.

Yorumlar (0)

Sonuc bulunamadi