senior-prompt-engineer

Use when the user asks to optimize prompts, design prompt templates, evaluate LLM outputs with an eval set, measure RAG retrieval quality, validate agent/tool configurations, analyze token usage, or design structured-output contracts. Covers eval-driven prompt iteration, RAG metrics (relevance, faithfulness, coverage), agent workflow validation, and token/cost budgeting — all model-agnostic, with three stdlib Python tools.

Quality

96%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

—

The risk profile of this skill

Quality

Content

92%Weight 40%Scale 1-5

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

The body is a tight, highly actionable reference: exact stdlib CLI invocations, five eval-gated workflows with explicit validation/feedback loops, and a well-signaled References table pointing to real one-level-deep files. Its only weakness is slight repetition of the model-agnostic framing.

Suggestions

Consolidate the model-agnostic/pricing rationale (currently restated in the intro, Operating Rule 5, and each tool note) into a single statement to tighten conciseness.

Consider moving the per-symptom fix table in the Prompt Optimization workflow into references/ to keep the core workflow even leaner.

The RAG Evaluator section says formats are 'printed in --help' — a one-line inline shape example would save a round-trip to --help.

Dimension	Reasoning	Score
Conciseness	Dense and assumes competence throughout — exact CLI flags, a leverage table, pointed operating rules — with no padding about basics Claude already knows; the only mild over-explanation is the model-agnostic framing repeated across the intro and tool notes, keeping it just below the lean score-5 anchor.	4 / 5
Actionability	Every tool section gives copy-paste-ready bash commands with real flags (verified against prompt_optimizer.py's argparse: --analyze/--tokens/--optimize/--extract-examples/--compare/--json/--output), and the eval-gate workflow includes an executable Python snippet, matching the fully-executable score-5 anchor.	5 / 5
Workflow Clarity	All five workflows are numbered with explicit validation gates and feedback loops — 'Eval gate (must pass before shipping)', 'every metric must be ≥ baseline; any regression blocks the change', 'must exit with VALIDATION PASSED' — directly matching the clear-sequence-with-validation score-5 anchor.	5 / 5
Progressive Disclosure	Well-organized sections with a References table that signals one-level-deep pointers (all three reference files verified present) plus a 'Load when user asks about' navigation column, matching the well-signaled single-level reference score-5 anchor.	5 / 5
	Total	19 / 20 Passed

Description

100%Weight 40%Scale 1-5

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

The description is a strong, third-person trigger description that names concrete actions, provides explicit 'Use when...' guidance, and carves out a distinct niche. It hits the top anchor on every dimension with no verbosity or over-claiming.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'optimize prompts, design prompt templates, evaluate LLM outputs with an eval set, measure RAG retrieval quality, validate agent/tool configurations, analyze token usage, or design structured-output contracts' — covering the full surface comprehensively, matching the score-5 anchor.	5 / 5
Completeness	Explicitly answers both what (the enumerated capabilities plus 'Covers eval-driven prompt iteration, RAG metrics... agent workflow validation, and token/cost budgeting') and when ('Use when the user asks to...') with concrete trigger phrases, matching the score-5 anchor.	5 / 5
Trigger Term Quality	The 'Use when the user asks to...' clause enumerates natural trigger phrases users would actually say (optimize prompts, measure RAG retrieval quality, design structured-output contracts) with strong synonym coverage; the file-extension criterion is N/A for this non-file skill, so comprehensive natural-term coverage holds at 5.	5 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (eval-driven prompt/RAG/agent-config validation) with distinct triggers and minimal conflict risk; the body's 'Related Skills' section further sharpens the boundary against sibling skills.	5 / 5
	Total	20 / 20 Passed

Validation

100%

Warnings & errors only

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: alirezarezvani/claude-skills
Path: engineering-team/skills/senior-prompt-engineer/SKILL.md
Commit: 19392f7

Reviewed: ~8 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.