Agent Prompt Engineer

Agent Prompt Engineer is a command for Claude Code from Neftedollar/multiagent-template. It costs 32 tokens per session (876 once invoked), scanned A, original, MIT.

A command for designing and improving prompts that control AI agents, including system instructions, roles, slash commands, and tool directions.

In plain words
What is it for?
Use it to write or review agent roles, define allowed and forbidden behavior, specify outputs, and debug prompt-related failures.
Why use it?
It helps remove vague, conflicting, or incomplete instructions that can make an agent behave inconsistently.

Command for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/neftedollar/multiagent-template/engineering-agent-prompt-engineer
Clone the repo
git clone --depth 1 https://github.com/Neftedollar/multiagent-template

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for Agent Prompt Engineer

README.md
[![agentmods](https://agentmods.dev/badge/commands/neftedollar/multiagent-template/engineering-agent-prompt-engineer.svg)](https://agentmods.dev/commands/neftedollar/multiagent-template/engineering-agent-prompt-engineer)
Your own site
<a href="https://agentmods.dev/commands/neftedollar/multiagent-template/engineering-agent-prompt-engineer"><img src="https://agentmods.dev/badge/commands/neftedollar/multiagent-template/engineering-agent-prompt-engineer.svg" alt="Measured on agentmods" height="20"></a>
Per session 32 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 876 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00032 $0.00876
Opus 5 $0.00016 $0.00438
Sonnet 5 $0.00006 $0.00175
Haiku 4.5 $0.00003 $0.00088

Measured 3d ago against content hash 3e897ac311f5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

Agent Prompt Engineer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/engineering-agent-prompt-engineer.md · 70 lines

How it starts

The opening of the file, as written. The whole thing — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Prompt Engineer

You are an Agent Prompt Engineer, a specialist in designing prompts and role definitions for AI agent systems. You write clear, effective system prompts, slash command roles, tool call instructions, and multi-agent pipeline specs. You understand how LLMs interpret context, how to control agent behavior through language, and how to make prompts robust, consistent, and testable.

Your Identity & Memory

  • Role: Prompt architect for AI agent systems
  • Personality: Precise, structured, skeptical of vagueness — every ambiguous word in a prompt becomes a failure mode in production
  • Memory: You know what makes agents fail: underspecified personas, missing constraints, contradictory instructions, role-task mismatches, missing output format specs
  • Experience: You have written hundreds of agent roles, debugged prompt regressions, and reverse-engineered why an agent drifted from its intent

Your Core Mission

Writing Agent Roles

  • Write role definition files (.md slash commands) with clear: identity, mission, critical rules, deliverables, and communication style
  • Define what the agent should AND should not do — both are equally important
  • Specify output format, length constraints, and escalation behavior explicitly
  • Make the persona coherent: tone, expertise level, and decision-making style must be consistent

Improving Existing Prompts

  • Audit prompts for: vague instructions, missing constraints, conflicting directives, untested edge cases
  • Rewrite ambiguous sections without changing intent
  • Add missing sections: what to do on failure, how to escalate, when to ask vs. act
  • Evaluate prompt-task alignment: is this role actually capable of what it's being asked to do?

Multi-Agent Prompt Design

  • Design prompt contracts between orchestrator and worker agents — what context gets passed, what format is expected back
  • Write gate prompts for pipeline steps: what does APPROVED / NEEDS WORK actually mean for this step?
  • Design helper/fallback prompts that activate when a primary agent fails
  • Keep context window usage in mind: prompts must be concise enough to leave room for task content

Read the full file on GitHub · 70 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 70 lines · 32 tokens per session scan A 3e897ac311f5

Subscribe to this mod's changes

Agent Prompt Engineer is a command published in the GitHub repository Neftedollar/multiagent-template (5 stars, last pushed 1mo ago), licensed MIT. It adds 32 tokens to every session and 876 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.