Skip to content
Back to skills

2478 Privacy 75152c5b

BSecurity

**Hardstop** is a local-only Claude Code plugin. This document describes what data it collects and how it's handled.

  • 9 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 11, 2026
ai-agentsapi

Works with

  • claude code
  • cli
  • api

Security analysis

B85/100
  • highPerforms destructive filesystem operations

Pro shows the line behind each finding and how to fix it

Scanned October 11, 2026

npx -y skills add tools-only/X-Skills --skill 2478-privacy_75152c5b --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of 2478 Privacy 75152c5b?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for 2478 Privacy 75152c5b
[![Security: B — Skills Directory](https://www.skillsdirectory.com/api/skills/tools-only-2478-privacy-75152c5b/badge)](https://www.skillsdirectory.com/skills/tools-only-2478-privacy-75152c5b)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

SKILL.md
# Privacy Policy

**Hardstop** is a local-only Claude Code plugin. This document describes what data it collects and how it's handled.

## Data Collection

### What Hardstop Collects

Hardstop only processes and stores data locally on your machine:

| Data | Purpose | Storage |
|------|---------|---------|
| Command text | Safety analysis | Logged to `~/.hardstop/audit.log` |
| Working directory | Context for analysis | Logged to audit.log |
| Verdict (ALLOW/BLOCK) | Decision record | Logged to audit.log |
| Timestamp | Audit trail | Logged to audit.log |

### What Hardstop Does NOT Collect

- No personal information
- No conversation history
- No API keys or credentials
- No telemetry or analytics
- No data sent to external servers

## Data Storage

All data is stored locally in `~/.hardstop/`:

```
~/.hardstop/
├── state.json    # Plugin enabled/disabled state
├── skip_next     # Temporary one-time bypass flag
└── audit.log     # Command decision log (JSON-lines)
```

## Data Transmission

**Hardstop sends no data to external servers.**

The only network activity is the optional LLM analysis layer, which uses the local Claude CLI (`claude --print`). This runs within your existing Claude subscription and is subject to Anthropic's privacy policy.

## Data Retention

- `audit.log` grows over time as commands are analyzed
- You can delete `~/.hardstop/` at any time to remove all stored data
- No automatic data expiration or cleanup

## Third-Party Services

Hardstop uses:
- **Claude CLI** (optional) — For LLM-based semantic analysis of edge cases
- No other third-party services

## Your Rights

You can:
- View all stored data: `cat ~/.hardstop/audit.log`
- Delete all data: `rm -rf ~/.hardstop/`
- Disable logging: Not currently supported (audit log is always written)

## Changes to This Policy

Updates will be documented in the repository changelog.

## Contact

For privacy questions: contact@clarity-gate.org

---

*Last updated: 2025-01-17*

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.