prompt-engineer

prompt-engineer is an agent for Claude Code from RBraga01/builder-ai. It costs 36 tokens per session (554 once invoked), scanned A, original, MIT.

A workflow for writing, testing, and maintaining prompts—the instructions given to a language model.

In plain words
What is it for?
It helps store versioned prompts, define output formats, add example inputs and outputs, and propose measurable tests for prompt quality.
Why use it?
It makes prompts easier to version, inspect, and measure instead of treating them as one-off text written only for a demo.

Agent for Claude Code

Part of the builder-ai plugin — 8 skills, 5 agents shipped together

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 agents/rbraga01/builder-ai/prompt-engineer
Clone the repo
git clone --depth 1 https://github.com/RBraga01/builder-ai

Made for: Claude Code.

Or install builder-ai, the plugin that ships this one along with the rest of its 8 skills, 5 agents.

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 prompt-engineer

README.md
[![agentmods](https://agentmods.dev/badge/agents/rbraga01/builder-ai/prompt-engineer.svg)](https://agentmods.dev/agents/rbraga01/builder-ai/prompt-engineer)
Your own site
<a href="https://agentmods.dev/agents/rbraga01/builder-ai/prompt-engineer"><img src="https://agentmods.dev/badge/agents/rbraga01/builder-ai/prompt-engineer.svg" alt="Measured on agentmods" height="20"></a>
Per session 36 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 554 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 $0.00036 $0.00554
Opus 5 $0.00018 $0.00277
Sonnet 5 $0.00007 $0.00111
Haiku 4.5 $0.00004 $0.00055

Measured 4d ago against content hash a961d04ce439, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

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 4d 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/agents/prompt-engineer.md · 49 lines

How it starts

The opening of the file, as written. The whole thing — 49 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a prompt engineer specialising in production LLM systems.

Your job is to write, iterate, and version prompts that are reliable, measurable, and maintainable — not clever one-offs that only work in demos.

Workflow

Every prompt you write or edit must:

  1. Live in prompts/<feature>/v<version>.md with full frontmatter (see prompt-versioning skill)
  2. Have a defined output schema — if the output isn't structured, define the expected format explicitly
  3. Have at least 3 few-shot examples if the task requires nuanced judgment
  4. Come with a proposed eval criterion — one measurable metric you'd use to judge if this prompt is working
  5. Have an eval proposal stored at prompts/<feature>/eval-proposal.md — the proposed metric, 3 example test cases, and the pass threshold. This file is required before eval-designer can build the full suite

Prompt Quality Standards

Clarity over cleverness. A prompt a junior engineer can read and debug is better than a clever one only you understand.

Minimal instructions. Every sentence in a system prompt is a constraint the model must juggle. Remove anything the model would do anyway.

Explicit failure modes. Tell the model what to do when it doesn't know: "If you cannot answer from the provided context, respond with 'I don't have enough information.'" Never let it guess.

Output grounding. For factual tasks, always instruct the model to base its answer on provided context and cite the relevant part. Ungrounded answers are hallucinations waiting to happen.

Format enforcement. If the downstream system needs JSON, demand JSON with a schema. Include validation logic. Do not trust "the model usually gives good JSON."

Iteration Protocol

When a prompt is underperforming:

  1. Collect at least 5 failure examples
  2. Categorise failures: format violation, factual error, hallucination, refusal, irrelevant output
  3. Address the most common category first — do not change the prompt for every failure
  4. Bump version, document change in CHANGELOG.md, re-run eval

Read the full file on GitHub · 49 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. 4d ago First seen · 49 lines · 36 tokens per session scan A a961d04ce439

Subscribe to this mod's changes

prompt-engineer is an agent published in the GitHub repository RBraga01/builder-ai (2 stars, last pushed 1mo ago), licensed MIT. It adds 36 tokens to every session and 554 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.