DeepEval Skills
Evaluation, tracing and OpenTelemetry export for LLM apps, split into three non-overlapping skills.
Install
OfficialClaude Code: /plugin marketplace add anthropics/claude-plugins-official → /plugin install deepeval@claude-plugins-official
Bir LLM uygulamasına değerlendirme, izleme ya da OpenTelemetry dışa aktarımı eklenirken
- Author
- Confident AI
- License
- Apache-2.0
Three skills, and the interesting part is the boundary between them. One builds evaluation suites: goldens, datasets, pytest runs, metrics, and the loop where failures drive prompt and retrieval changes. One instruments an application with DeepEval's own tracing and the framework integrations that ship with it. One covers plain OpenTelemetry export for teams that do not want the SDK. Each frontmatter states what it must not handle and names the sibling that should, which is a cheap answer to the usual problem of three adjacent skills all firing on the word eval.
Similar skills
LLM Application Dev
SkillBundles the RAG, vector-index and prompting skills an LLM app needs, rebuilt around LangGraph.
PostHog
SkillDozens of narrow skills for PostHog analytics, flags and experiments — and a channel back the other way.
PluginEval
SkillScores a Claude Code skill on ten dimensions across static checks, an LLM judge and Monte Carlo runs.
Comments(0)
Sign in to comment