What is goal-engineering AGENTS.md?

Repository instructions for goal-engineering: a structured way to define work objectives, completion conditions, and verification. They describe goal patterns, supporting tools, documentation, and rules for recording outcomes.

Use them when starting or auditing a goal, checking whether work is complete, estimating effort, running repository validation, or recording a goal run.

They help keep coding work focused and prevent claiming a task is finished before its checks pass. They also define boundaries around unrelated or unsafe changes.

goal-engineering AGENTS.md — Instructions file for Codex and OpenCode by cobusgreyling: Repository instructions for goal-engineering: a stru

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 instructions/cobusgreyling/goal-engineering/agents-md

Clone the repo

git clone --depth 1 https://github.com/cobusgreyling/goal-engineering

Made for: Codex, OpenCode.

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/cobusgreyling/goal-engineering/agents-md.svg)](https://agentmods.dev/instructions/cobusgreyling/goal-engineering/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/cobusgreyling/goal-engineering/agents-md"><img src="https://agentmods.dev/badge/instructions/cobusgreyling/goal-engineering/agents-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 401 This file is loaded in full into every session.

When invoked 401 The same file — it is already loaded in full.

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.

Model	Per session	Once invoked
Fable 5	$0.00401	$0.00401
Opus 5	$0.00200	$0.00200
Sonnet 5	$0.00080	$0.00080
Haiku 4.5	$0.00040	$0.00040

Measured 4d ago against content hash 5eaac391ef44, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

goal-engineering AGENTS.md 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.

AGENTS.md · 49 lines

What it actually says

AGENTS.md — Goal Engineering Reference

Project Purpose

Canonical public reference for Grok Build CLI /goal and update_goal. Docs-first; CLIs in tools/.

Path	Contents
`patterns/`	Six goal patterns + `registry.yaml`
`starters/`	`minimal-goal`, `tests-green`, `fix-bug`
`skills/`	`goal-verifier`, `goal-scoper`, `goal-completion-check`
`.grok/skills/`	Grok-native copy of skills (dogfood)
`tools/goal-audit`	G0–G3 readiness CLI
`tools/goal-init`	Scaffold patterns into projects
`tools/goal-cost`	Turn/token estimates
`docs/`	API, safety, anti-patterns, multi-goal
`GOAL.md`	Active repo objective + done conditions

Verification

npm test
npm run validate

Before claiming work done: run tests; use goal-verifier mindset on doc accuracy.

Goal Discipline

Mirror active /goal objectives in GOAL.md
Never update_goal(completed: true) without verifier PASS
One active goal per session
Log outcomes in goal-run-log.md

Deny List

npm publish without version bump + release tag
GitHub org settings or unrelated user repos
Changing loop-engineering from this repo (cross-link only)

Code Style

TypeScript ESM in tools
Markdown: clear headings, copy-paste /goal examples
Match tone of loop-engineering

Commits

Conventional commits: docs:, feat(goal-audit):, feat(goal-init):, chore:.

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.

4d ago First seen · 49 lines · 401 tokens per session scan A 5eaac391ef44

Subscribe to this mod's changes

goal-engineering AGENTS.md is an instructions file published in the GitHub repository cobusgreyling/goal-engineering (40 stars, last pushed 2mo ago), licensed MIT. It adds 401 tokens to every session, about $0.0020 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 2026-08-30.

Planning ai-agents automation goal-engineering grok grok-build verifier-skills

grok-keysmith AGENTS.md

AGENTS.md instructions for Jia-Ethan/grok-keysmith, covering repository agent guide, scope, release and external state, change boundaries and documentation rules.

Jia-Ethan/grok-keysmith · 822 tokens

pi-xai-oauth AGENTS.md

AGENTS.md instructions for BlockedPath/pi-xai-oauth, covering agents.md — ai agent operations manual for pi-xai-oauth, project overview, key commands (exact, copy-paste ready), architecture & boundaries (must / must not) and file structure & wayfinding.

BlockedPath/pi-xai-oauth · 2,961 tokens

goal-engineering AGENTS.md