Write, optimize, and debug LLM prompts — system prompts, task prompts, and prompt templates. Trigger when: the user wants to write or improve a system prompt; asks 'how do I prompt Claude/GPT/Gemini to [X]'; a prompt keeps giving bad or inconsistent outputs; wants to build a prompt template; asks about few-shot, chain-of-thought, or role-based prompting; wants to reduce hallucinations or improve AI output quality; or asks to evaluate or critique a prompt. Key capabilities: role assignment…
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Prompt Engineering Skill
## Core Techniques
### 1. Be Specific About Output Format
Always specify: format (JSON, markdown, bullet list), length, tone, and structure.
```
Bad: "Summarize this article"
Good: "Summarize this article in 3 bullet points, each under 20 words, focusing on business impact"
```
### 2. Give the Model a Role
```
"You are a senior software engineer reviewing code for production readiness.
Your goal is to find bugs, security issues, and performance problems."
```
### 3. Use Examples (Few-Shot)
Include 2-3 examples of input → output pairs before the actual task.
### 4. Chain of Thought
For reasoning tasks, add: *"Think step by step before giving your final answer."*
### 5. XML Tags for Structure
```xml
<context>Background information here</context>
<task>What you want done</task>
<constraints>Rules to follow</constraints>
<output_format>Expected format</output_format>
```
### 6. Positive Instructions Over Negative
```
Bad: "Don't write long responses"
Good: "Keep responses under 100 words"
```
## Concrete Examples
### Before/After Prompt Improvement
**Weak:** "Summarize this article"
…
The line at the top of this page pins the label git:20260508.65c1b0b, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
GET https://markdownregistry.com/api/v1/artifacts/art_da6i5epvztx5lh5e
GET https://markdownregistry.com/api/v1/resolve?ref=tekkiiiii/the-agency/prompt-engineering
GET https://markdownregistry.com/api/v1/blob/86e4da5dc4b30abd746a064d5ec287a6baff95887d8d9093683299904c578633
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
tekkiiiii/the-agency · agents-archive/dept-sunset-2026-10-08/skills/dept-resume/SKILL.md · Read dept-state.md and spawn a Dept Head with a lean briefing. Department-operations parallel of /pd-resume.
tekkiiiii/the-agency · agents-archive/dept-sunset-2026-10-08/skills/dept-save-state/SKILL.md · Write dept-state.md and member-roster.md at end of department session. Department-operations parallel of /save-state.
tekkiiiii/the-agency · agents-archive/dept-sunset-2026-10-08/skills/dept-status/SKILL.md · Read-only department status digest. No spawns. Shows active coords, pipelines, open issues, and incoming tasks.
tekkiiiii/the-agency · agents-archive/dept-sunset-2026-10-08/skills/room-manager-digest/SKILL.md · Use on /room-manager-digest, a 12h cron, or when a dept head wants a fresh report. Reads each dept room's rolling.md…
tekkiiiii/the-agency · agents-archive/dept-sunset-2026-10-08/skills/room-manager/SKILL.md · Use on /room-manager, every 10 minutes as a background loop, or when new room activity is suspected (handoff file…
tekkiiiii/the-agency · agents-archive/rev001-curator-2026-10-12/skills/obsidian-vault/SKILL.md · Use to forward decisions, project learnings, skill creations and generated assets to the Obsidian vault, and to read it…
tekkiiiii/the-agency · scripts/tests/fixtures/jev/router-skills/content-polish/SKILL.md · Polish drafted content to publish-ready in one pass; English and Vietnamese.
tekkiiiii/the-agency · scripts/tests/fixtures/jev/router-skills/frontend/SKILL.md · Use when building frontend UIs: "build a landing page", "make a dashboard", "write a React component". Not for…
giuseppe-trisciuoglio/developer-kit · plugins/developer-kit-ai/skills/prompt-engineering/SKILL.md · Provides workflows to write, debug, and optimize prompts for LLMs, including few-shot example selection…
ericrisco/rsc-harness · skills/prompt-engineering/SKILL.md · Use when one prompt must give the same right answer across reruns, models, and pasted-in hostile input: forcing a fixed…
codealive-ai/ai-driven-development · skills/prompt-engineering/SKILL.md · Universal prompt engineering techniques for any LLM. Use when crafting, optimizing, or reviewing prompts for AI models…
doodledood/manifest-dev · claude-plugins/manifest-dev-tools/skills/prompt-engineering/SKILL.md · Create, update, review, or discuss an LLM prompt — a system prompt, a skill, or an agent. Use when writing or improving…
nahid-sparktales/agent-dispatcher · skills/ai/prompt-engineering/SKILL.md · Write or revise a prompt so it holds up — output contract, instruction placement, examples that earn their place, an…
vodailocz/kilo-kit-mcp · skills/engineering/prompt-engineering/SKILL.md · Use when designing, optimizing, testing, or deploying robust prompt systems for AI agents. This skill provides…
