ng-drift-check

ng-drift-check is a command for coding agents from FlyFission/nuclear-grade-context-engineering. It costs 0 tokens per session (875 once invoked), scanned A, original, MIT.

A command that checks whether ongoing agent work still matches its written mission and scope.

In plain words
What is it for?
It is for deciding whether to continue, refocus, ask for help, or stop when work no longer clearly serves the original goal.
Why use it?
It helps catch scope creep, repeated failed attempts, drifting priorities, and lowered standards during long tasks.

Command

Part of the nuclear-grade plugin — 24 skills, 28 commands, 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 commands/flyfission/nuclear-grade-context-engineering/ng-drift-check
Clone the repo
git clone --depth 1 https://github.com/FlyFission/nuclear-grade-context-engineering

Or install nuclear-grade, the plugin that ships this one along with the rest of its 24 skills, 28 commands, 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 ng-drift-check

README.md
[![agentmods](https://agentmods.dev/badge/commands/flyfission/nuclear-grade-context-engineering/ng-drift-check.svg)](https://agentmods.dev/commands/flyfission/nuclear-grade-context-engineering/ng-drift-check)
Your own site
<a href="https://agentmods.dev/commands/flyfission/nuclear-grade-context-engineering/ng-drift-check"><img src="https://agentmods.dev/badge/commands/flyfission/nuclear-grade-context-engineering/ng-drift-check.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 875 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 $0.00000 $0.00875
Opus 5 $0.00000 $0.00438
Sonnet 5 $0.00000 $0.00175
Haiku 4.5 $0.00000 $0.00088

Measured 5d ago against content hash 91eae7c12fa9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ng-drift-check 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 5d 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.

commands/ng-drift-check.md · 69 lines

How it starts

The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.

ng-drift-check

Portable command prompt generated from skills/staying-on-mission/SKILL.md. Edit the skill, then run python tools/ng.py gen-commands; do not edit this file by hand.

Tests the current work against a written, lasting mission anchor and forces one choice, re-anchor, escalate, or stop. Use when an agent keeps finishing tasks but the work drifts from the goal, scope creeps, the same action repeats in a loop, or standards slip one small step at a time. Do not use for a tiny edit with an obvious goal, or while you are still containing a live incident.

Use when

  • A long work session has many steps, and the current action is hard to tie back to the original goal.
  • Scope is growing. New files, features, or layers show up that no one asked for.
  • The same action, file, or fix keeps getting retried with no progress. That is a loop.
  • The agent is spending a lot (high token use) but you can see little real progress.
  • A reviewer or agent is working from memory of the goal instead of a written anchor.
  • A non-goal is about to be crossed, or a standard is about to be loosened "just this once."
  • Context was reset, shrunk, or handed off, so the goal must be set again before work goes on.

Do not use when

  • A tiny quick edit with an obvious goal and no risk of scope growth.
  • A live incident you have to contain before you stop to reflect.
  • The user is asking for a formal guarantee, a certification, or regulatory approval.

Inputs

  • The mission anchor: .nuclear/mission.md, the ## Mission anchor in risk.md, or the issue or pull request it came from.
  • The repo charter (.nuclear/charter.md) when there is one: the lasting principles the work must not break.
  • The current action and what was tried recently, including how many times.
  • The files the change touches, the diff so far, and the stated non-goals.

Prompt text

Run a Nuclear-grade mission-drift check on the current work.

Inputs:
- mission anchor (objective, success criteria, non-goals):
- charter principles in play:
- current action:
- recent attempts at this objective (how many, what variants):
- affected files / diff so far:

Do this:
- Restate the goal from the written record, not from memory.
- Zoom out one layer; judge at the level of the goal and the architecture, not the detail.
- Test the current action against the success criteria and the non-goals.
- Decide whether the action serves the mission, or a smaller local goal that has quietly replaced it.
- Check the loop: if the same goal has failed 3 times, stop trying the next variant.
- Check for slipping standards against the charter and any countable tripwires.

Return one decision:
- RE-ANCHOR: the action serves the mission; restate the goal and continue.
- ESCALATE: a non-goal or a standard must be crossed for a defensible reason; include a justification row (what is crossed, why, and why no simpler path exists).
- STOP: the action serves a smaller local goal, or the justification does not hold.

Also return the updated goal text, so the decision survives the next context reset.
Do not imply formal assurance, compliance, certification, safety, security, or regulatory adequacy.

Read the full file on GitHub · 69 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. 5d ago First seen · 69 lines · 0 tokens per session scan A 91eae7c12fa9

Subscribe to this mod's changes

ng-drift-check is a command published in the GitHub repository FlyFission/nuclear-grade-context-engineering (33 stars, last pushed 26d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 875 tokens. 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.