prompt-injection-defense

SKILLWorkflowcommunity
v0.0.0GoldenWing-360MITUpdated 13d agoSource →

Contain direct and indirect prompt injection in LLM-integrated applications. Covers source-of-trust tagging, tool-use confirmation after untrusted input, output validation, markdown-image exfiltration prevention, and context-window hygiene. Invoke when building any app where untrusted text reaches a

Community-submitted skill. Not yet reviewed by the Forge team. Full prompt content may not be available.Request review →
15Repo stars
1Clients
1Formats
13d agoLast update
Skill
AuthorGoldenWing-360
Version0.0.0
LicenseMIT
CategoryWorkflow
Formatsskill.md
PromptNot published
Compatibility
Claude✓ Supported
Cursor
Copilot
ChatGPT
Gemini
About

Contain direct and indirect prompt injection in LLM-integrated applications. Covers source-of-trust tagging, tool-use confirmation after untrusted input, output validation, markdown-image exfiltration prevention, and context-window hygiene. Invoke when building any app where untrusted text reaches an LLM, when the LLM has tools that act on real systems, or after a suspected injection incident.

Keywords
skillclaude