Guides

Prompt Injection Defense Toolkit for Claude Code Users

Detect injected instructions in fetched content, audit tool call chains, and run agentic workflows safely — the prompt injection defense guide Anthropic hasn't written yet.

$ tar -tzf prompt-injection-defense-toolkit-for-claude-code-users.tar.gz
  • Prompt Injection Defense Guide — 7-chapter reference covering threat modeling, detection, CLAUDE.md hardening, MCP trust policy, tool-call auditing, and incident response
  • Hardened CLAUDE.md Template — drop-in configuration with injection-resistant pre-wiring; copy to any project, fill in four fields, done
  • MCP Trust Policy Reference — four-tier risk model, 20-item per-server vetting checklist, and .mcp.json allowlist patterns
  • Injection Detection Heuristics — 25 red-flag patterns in fetched content, from naive override phrases to obfuscated Unicode tricks and multi-turn persistence attacks
  • Tool-Call Audit Log Template — structured session log to track what Claude Code called, what it returned, and whether you reviewed it
  • Pre-Flight Checklist — 20-item pre-session checklist covering CLAUDE.md state, MCP server scope, and environment verification
$19

one-time purchase

Instant download after purchase
📧Download link sent to your email
🔄7-day download access
14-day money-back guarantee
View refund policy

## goes-well-with

From the same shelf

## not-ready-to-buy

Take the field notes instead

One practical write-up a week from the same workbench these kits come from — plus reader pricing when new kits ship.