exploring-llm-evaluations
exploring-llm-evaluations is an agent skill published by posthog, listed under review & testing. It lives in the PostHog/skills repository, which has 75 GitHub stars and is licensed under MIT; we last read it on 2026-10-18.
Investigate AI observability evaluations — `hog` (deterministic code-based), `llm_judge` (LLM-prompt-based), and `sentiment` (user-message sentiment). Find existing evaluations, inspect their configuration, run them against specific generations, query individual results, and set up scheduled reports on an evaluation. Use when the user asks to debug why an evaluation is failing, surface common failure modes, compare results across filters, dry-run a Hog evaluator, prototype a new LLM-judge prompt, inspect sentiment classifications, or manage the evaluation lifecycle.
Install
skills is a generic installer for any repository of Claude agent skills -- it reads PostHog/skills’s own structure rather than assuming one, the same way this page reads the skill’s own SKILL.md rather than assuming a format.
npx skills add PostHog/skills
Maintenance
Investigate AI observability evaluations — `hog` (deterministic code-based), `llm_judge` (LLM-prompt-based), and `sentiment` (user-message sentiment). Find existing evaluations, inspect their configuration, run them against specific generations, query individual results, and set up scheduled reports on an evaluation. Use when the user asks to debug why an evaluation is failing, surface common failure modes, compare results across filters, dry-run a Hog evaluator, prototype a new LLM-judge prompt, inspect sentiment classifications, or manage the evaluation lifecycle.
exploring-llm-evaluations compared with similar skills
| Skill | Publisher | What it does | Repo stars | Last update |
|---|---|---|---|---|
| exploring-llm-evaluations (this page) | posthog | Investigate AI observability evaluations — `hog` (deterministic code-based), `llm_judge` (LLM-prompt-based),… | 75 | 2026-10-25 |
| signals-scout-web-vitals(same repository) | posthog | Signals scout for Core Web Vitals (`$web_vitals`). Watches each page's p75 LCP / INP / CLS / FCP against Goog… | 75 | 2026-10-25 |
| code-review | mattpocock | Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does t… | 303,549 | 2026-10-22 |
| scaffold-exercises | mattpocock | Create exercise directory structures with sections, problems, solutions, and explainers that pass linting. Us… | 303,549 | 2026-10-22 |
| tdd | mattpocock | Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-gree… | 303,549 | 2026-10-22 |
| webapp-testing | anthropic | Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend… | 183,845 | 2026-10-09 |
| browser-testing-with-devtools | addyosmani | Tests in real browsers via Chrome DevTools MCP. Use when building or debugging anything that runs in a browse… | 110,611 | 2026-10-13 |
Skills from the same repository first, then others for this job by repository stars. Figures are from each repository when we last read it.
Common questions about exploring-llm-evaluations
- What does exploring-llm-evaluations do?
- Investigate AI observability evaluations — `hog` (deterministic code-based), `llm_judge` (LLM-prompt-based), and `sentiment` (user-message sentiment). Find existing evaluations, inspect their configuration, run them against specific generations, query individual results, and set up scheduled reports on an evaluation. Use when the user asks to debug why an evaluation is failing, surface common failure modes, compare results across filters, dry-run a Hog evaluator, prototype a new LLM-judge prompt, inspect sentiment classifications, or manage the evaluation lifecycle.
- How do I install exploring-llm-evaluations?
- Run npx skills add PostHog/skills. The skills installer reads the repository’s own structure, so it works for any repository of agent skills.
- Who publishes exploring-llm-evaluations?
- posthog.
- Is exploring-llm-evaluations maintained, and what licence does it use?
- Its repository has 75 GitHub stars and its state in our directory is listed. Its last commit we recorded was on 2026-10-25. The repository declares the MIT licence.
More agent skills for review & testing
- browser-testing-with-devtoolsby addyosmani
- code-review-and-qualityby addyosmani
- code-simplificationby addyosmani
- doubt-driven-developmentby addyosmani
- vitestby antfu
- webapp-testingby anthropic