evaluate-llm-feature

Name: evaluate-llm-feature
Author: event4u-app

SKILLWorkflowcommunity

v0.0.0event4u-appMITUpdated 1d agoSource →

Black-box evaluation of a shipped LLM feature — adversarial probes for hallucination, prompt-injection, and cost-runaway vs stated expectations. Not RAG/embedding. Triggers 'review my chatbot'.

Community-submitted skill. Not yet reviewed by the Forge team. Full prompt content may not be available.Request review →

7Repo stars

1Clients

1Formats

1d agoLast update

Skill

Authorevent4u-app

Version0.0.0

LicenseMIT

CategoryWorkflow

Formatsskill.md

PromptNot published

Compatibility

Claude✓ Supported

Cursor—

Copilot—

ChatGPT—

Gemini—

About

Black-box evaluation of a shipped LLM feature — adversarial probes for hallucination, prompt-injection, and cost-runaway vs stated expectations. Not RAG/embedding. Triggers 'review my chatbot'.

Keywords

skillclaude