Back to catalog
Human reviewedDevelopment

Skill dossier

prompt-engineer

Source declaration

Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot learning, creating system prompts with personas and guardrails, building JSON/function-calling schemas, or developing prompt evaluation frameworks to measure and improve model performance.

The author's original description, preserved in English. It describes intent, not verified performance.

Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot learning, creating system prompts with personas and guardrails, building JSON/function-calling schemas, or developing prompt evaluation frameworks to measure and improve model performance.

developmentreferences

Research summary

Human reviewed

Skill Atlas editorial review · editorial-preview-1.0.0

Guides prompt design, testing and iteration for LLM applications, including few-shot and chain-of-thought patterns, structured output schemas and evaluation plans.

Review scope

Deterministic static analysis of collected SKILL.md and supporting text/code. Machine-assisted candidates and findings await human editorial review; no dynamic behavior or effect is assessed.

Observable capabilities

  • Defines task, success criteria, constraints and edge cases before writing a prompt.
  • Chooses a prompting pattern (zero-shot, few-shot, chain-of-thought) and writes the instructions.
  • Tests prompts on diverse inputs and changes one thing at a time when iterating.
  • Delivers the final prompt with test cases, usage settings and known limitations.

Limitations & scope

  • The skill sets an accuracy checkpoint of 80% but does not supply a test harness or datasets.
  • Model-specific behavior is delegated to reference files that this review did not exercise.
  • Results here are unverified; no prompt produced by the skill was run against a model.

Relevant for

  • Developers building LLM-backed features
  • Teams refactoring prompts for accuracy or token cost
  • Engineers designing evaluation suites for prompts

What ships with the skill

The file inventory comes from the collected source at the recorded commit. Paths and sizes are structural observations, not a measure of quality.

files
7
Markdown files
7
scripts
0
references
6
Size
106.7 KB
Structural complexity
Moderate

File inventory

PathTypeBytes
skills/prompt-engineer/references/context-management.mdReference7.4 KB
skills/prompt-engineer/references/evaluation-frameworks.mdReference23.2 KB
skills/prompt-engineer/references/prompt-optimization.mdReference17.2 KB
skills/prompt-engineer/references/prompt-patterns.mdReference15.8 KB
skills/prompt-engineer/references/structured-outputs.mdReference18.6 KB
skills/prompt-engineer/references/system-prompts.mdReference18.5 KB
skills/prompt-engineer/SKILL.mdInstruction5.9 KB

SKILL.md frontmatter

{
  "name": "prompt-engineer",
  "description": "Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot learning, creating system prompts with personas and guardrails, building JSON/function-calling schemas, or developing prompt evaluation frameworks to measure and improve model performance.",
  "license": "MIT",
  "metadata": {
    "author": "https://github.com/Jeffallan",
    "company": "https://synergetic.solutions",
    "version": "1.2.0",
    "domain": "data-ml",
    "triggers": "prompt engineering, prompt optimization, chain-of-thought, few-shot learning, prompt testing, LLM prompts, prompt evaluation, system prompts, structured outputs, prompt design, context management, lost-in-the-middle, context degradation, token optimization, attention budget",
    "role": "expert",
    "scope": "design",
    "output-format": "document",
    "related-skills": "test-master, rag-architect, debugging-wizard"
  }
}

A reading of the operating sequence

This workflow reconstructs the apparent sequence from source material. Unless controlled evaluation is explicitly recorded, it is provisional—not an observed trace.

  1. 01
    Input

    Understand requirements

    Define the task, success criteria, constraints and edge cases.

  2. 02
    Action

    Design initial prompt

    Choose zero-shot, few-shot or chain-of-thought and write clear instructions.

  3. 03
    Decision

    Check accuracy checkpoint

    If accuracy is below 80% on the test set, identify failure patterns before iterating.

  4. 04
    Output

    Deliver prompt package

    Provide the final prompt, test cases and results, usage settings, metrics and known limitations.

Potential risk surfaces

Static findings identify points to inspect before installation. They do not establish that a skill is safe, malicious, or effective.

Attention recommended

86 deterministic lexical finding(s) require review across the collected text. Findings can occur in documentation and do not prove execution or exploitability; binary-only files were not scanned.

Network accessObserved
File writesObserved
Credential useNot observed
Medium

Network access is mentioned

6 occurrences

A URL or network client appears in this source line. Review destinations, data sent, and trust boundaries; static text cannot establish actual network activity. This rule matched 6 location(s) in the collected text; 3 representative location(s) are cited. Matching text may be documentation, and the analyzer did not execute it.

Note

File modification is described or implemented

76 occurrences

This line contains a possible file-write or redirection operation. Check destination paths and overwrite behavior; the static review did not write any file. This rule matched 76 location(s) in the collected text; 3 representative location(s) are cited. Matching text may be documentation, and the analyzer did not execute it.

High

Prompt-override language is present

4 occurrences

Instruction-override language appears in supplied content. Treat it as untrusted text and assess whether it tries to supersede higher-priority instructions or expose hidden context. This rule matched 4 location(s) in the collected text; 3 representative location(s) are cited. Matching text may be documentation, and the analyzer did not execute it.

Editorial security note

Confirm prompts and examples you feed it contain no sensitive data, since the skill itself says not to hardcode such data.

Where context can be spent

Discovery, activation, and resource layers have different meanings. Exact values appear only when the official token-counting method was run with a named model.

Exact token counts unavailable

All token values, including per-file inventory, are null until the pipeline performs an official exact count with an explicit model.

No token measurement was recorded with the required official method. This dossier does not substitute character counts, file sizes, or estimates.

Controlled evaluation

No controlled evaluation

This skill has only the evidence identified above. No task success, token efficiency, latency, or quality advantage is claimed.

Not tested

Use with the source in view

Install command
git clone https://github.com/Jeffallan/claude-skills.git cd claude-skills # Copy this skill directory into one of: # ~/.claude/skills/prompt-engineer # .claude/skills/prompt-engineer
  1. Open the pinned source at https://github.com/Jeffallan/claude-skills/blob/67c1961b180dc92cf08983344627109e983e8f51/skills/prompt-engineer/SKILL.md.
  2. Review the license, prerequisites and all bundled resources.
  3. Install the skill into a personal or project-level Claude Code skills directory.
  4. Restart or refresh Claude Code so the skill can be discovered.
Verify installation

Check that SKILL.md for prompt-engineer is discoverable at the destination, then ask Claude Code to use it on a small test task.

Prerequisites

  • The skill sets an accuracy checkpoint of 80% but does not supply a test harness or datasets.
  • Model-specific behavior is delegated to reference files that this review did not exercise.
  • Results here are unverified; no prompt produced by the skill was run against a model.

Example prompts

Representative task

Use the prompt-engineer skill to refactor my customer-support ticket classifier prompt. Define success criteria, propose a few-shot version, and write a test set including edge cases such as empty and multilingual tickets.

Inspect before use

Read the prompt-engineer skill files at https://github.com/Jeffallan/claude-skills/blob/67c1961b180dc92cf08983344627109e983e8f51/skills/prompt-engineer/SKILL.md. Explain its workflow, required tools, permissions and the main risks before I decide whether to install it.

A decision, with its limits attached

Strengths

  • Requires evaluation on diverse inputs before deployment.
  • Asks for one change at a time when debugging.

Considerations

  • Needs your own test data and metrics to be useful.
  • Prompt behavior may not transfer between models.
Research conclusion

A disciplined prompt-engineering checklist that is only as good as the test cases you bring.

Traceable by construction

The dossier is a time-bound reading of a public artifact, not an endorsement. Use the source link to inspect the exact upstream material.

Jeffallanclaude-skillsskills/prompt-engineer
Source type
Community
Commit
67c1961b180dc92cf08983344627109e983e8f51
Collected
20 Nov 2026
Last checked
20 Nov 2026
Upstream license
MIT
Repository stars
12,464
Content basis

Deterministic lexical parsing, YAML frontmatter validation, file inventory, source-described workflow extraction, and heuristic provisional classification of the supplied files only. Generated by machine; no LLM, execution, dynamic testing, or human editorial approval.