Add jpi-guard and inject-guard-en to LLM Prompt Injection Prevention Security Tools

[DOKASUKA]

Thank you for submitting a Pull Request (PR) to the Cheat Sheet Series.

Summary

This PR adds two prompt injection detection API tools to the Security Tools section of the LLM Prompt Injection Prevention Cheat Sheet.

Why these tools are relevant:

The existing Security Tools section lists NeMo Guardrails and Garak as detection/prevention tools. Two additional tools address specific threat vectors documented in this cheat sheet that are not yet covered by existing tools:

• Full-width Unicode bypass attacks (documented in the "Encoding and Obfuscation Techniques" section): ａｂｃ → abc normalization is a real attack vector targeting Japanese-language LLM applications and any system accepting multilingual input. jpi-guard specifically handles this normalization before detection.
• Polite-language disguise attacks: Japanese grammar allows embedding imperatives inside polite phrasing that a standard English-pattern regex misses. This is an under-documented attack surface.
• Indirect injection from external content (documented in the "Remote/Indirect Prompt Injection" section): inject-guard-en provides an API endpoint for sanitizing content fetched from external sources before passing it to an LLM.

Both tools are production APIs with free tiers, lowering the barrier to adoption for developers reading this cheat sheet.

Checklist

• In case of a new Cheat Sheet, you have used the Cheat Sheet template. (N/A — this is a minor addition to an existing cheat sheet)
• All the markdown files do not raise any validation policy violation.
• All the markdown files follow these format rules.
• All your assets are stored in the assets folder. (N/A — no assets added)
• All the images used are in the PNG format. (N/A)
• Any references to websites have been formatted as [TEXT](URL)
• You verified/tested the effectiveness of your contribution — both APIs have been tested against the attack patterns documented in this cheat sheet (direct injection, encoding obfuscation, and indirect injection from external content).
• The CI build of your PR pass.

AI Tool Usage Disclosure (required for all PRs)

• I have used AI tools to generate the contents of this PR. I have verified the contents and I affirm the results. The LLM used is Claude Sonnet 4.6 and the prompt used was to draft the tool descriptions in a style consistent with the existing Security Tools section. The tool descriptions were reviewed and verified against the actual API behavior before submission.

Thank you for your consideration!