Prompt injection detection skill

# Prompt injection detection skill

Two-layer content safety for agent input and output. Use when (1) a user message attempts to override, ignore, or bypass previous instructions (prompt injection), (2) a user message references system prompts, hidden instructions, or internal configuration, (3) receiving messages from untrusted users in group chats or public channels, (4) generating responses that discuss violence, self-harm, sexual content, hate speech, or other sensitive topics, or (5) deploying agents in public-facing or multi-user environments where adversarial input is expected.

## Discovery Metadata
- Category: `coding`
- Framework: `ClawHub`
- Tags: `prompt`, `injection`, `content`, `input`, `user`, `message`, `instructions`, `detection`, `skill`, `two-layer`

## Agent Execution Policy

This listing is **discovery metadata only**. Canonical instructions are maintained by ClawHub.

### Before Executing Actions

1. **Fetch canonical instructions** from: https://clawhub.ai/skill/detect-injection
2. **Parse the skill page** for setup, usage, and input/output requirements.
3. **Only proceed** after successfully loading and understanding the full instructions.

### If Fetch Fails

- Return `instruction_unavailable` with reason.
- Do **not** attempt to infer or improvise execution steps from this metadata alone.

## Source

- ClawHub listing: https://clawhub.ai/skill/detect-injection