CtrlK
BlogDocsLog inGet started
Tessl Logo

testland/promptfoo-evaluation

Authors and runs Promptfoo evals for LLM prompts and RAG pipelines - wires `promptfooconfig.yaml` providers + prompts + tests + assertions (deterministic `equals` / `contains` / `is-json` / `regex`, semantic `similar`, model-graded `llm-rubric` / `factuality` / `g-eval`, performance `latency` / `cost`, custom `javascript` / `python`), runs `npx promptfoo eval`, views HTML report via `promptfoo view`, and integrates CI for regression gating. Use when the user runs Promptfoo, asks about prompt regression suites, or needs an eval-driven workflow for LLM-backed features.

73

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

Overview
Quality
Evals
Security
Files

Quality

Content

80%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-organized, highly actionable reference with lean tables and executable configs, weakened only by an implicit (rather than explicit) validation feedback loop and inline reference material that could be split out.

Suggestions

Add an explicit validate→fix→retry checkpoint (e.g., 'After promptfoo eval, review failing assertions in the HTML report, fix prompts/assertions, then re-run eval until green') to lift workflow clarity.

Move the detailed assertion catalog (Step 4) and GitHub Action block (Step 8) into reference files (e.g., ASSERTIONS.md, CI.md) and link to them from the body to improve progressive disclosure.

Pin or call out version-sensitive items (e.g., the Node.js 20.20+/22.22+ requirement and pinned model snapshots) in a dedicated section so time-sensitive details don't clutter the core workflow.

DimensionReasoningScore

Conciseness

The body is dense tables and executable snippets that assume Claude's competence — it never explains what YAML, assertions, or embeddings are — so most tokens earn their place, matching the lean score-3 anchor.

3 / 3

Actionability

Provides copy-paste-ready commands ('npm install -g promptfoo', 'promptfoo eval') and complete YAML configs including the GitHub Action, matching the fully-executable score-3 anchor.

3 / 3

Workflow Clarity

Eight numbered steps give a clear sequence, but there is no explicit validate→fix→retry checkpoint in the authoring flow (eval/view/CI-gate are only implicit), fitting the score-2 'checkpoints missing or implicit' anchor rather than the score-3 explicit-validation anchor.

2 / 3

Progressive Disclosure

External doc links are one level deep and clearly signaled, but no bundle files exist and the assertion catalog and CI details are substantial inline content that could live in separate reference files, fitting the score-2 'content that should be separate is inline' anchor.

2 / 3

Total

10

/

12

Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, specific description that names concrete actions and provides explicit 'Use when' triggers tied to natural user language. It clearly distinguishes Promptfoo eval work from sister skills.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — 'Authors and runs Promptfoo evals', 'wires promptfooconfig.yaml providers + prompts + tests + assertions', 'runs npx promptfoo eval', 'views HTML report via promptfoo view', 'integrates CI for regression gating' — matching the score-3 anchor.

3 / 3

Completeness

Explicitly answers both what ('Authors and runs Promptfoo evals...') and when ('Use when the user runs Promptfoo, asks about prompt regression suites, or needs an eval-driven workflow...'), matching the score-3 anchor.

3 / 3

Trigger Term Quality

Natural user-facing terms are well covered: 'Promptfoo', 'prompt regression suites', 'eval-driven workflow for LLM-backed features', aligning with the score-3 anchor for good coverage of terms users would say.

3 / 3

Distinctiveness Conflict Risk

Occupies a clear Promptfoo-specific niche with distinct triggers tied to 'Promptfoo' and 'prompt regression suites', making it unlikely to fire for the wrong skill.

3 / 3

Total

12

/

12

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Reviewed

Table of Contents