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.
npx agentmods add commands/flyfission/nuclear-grade-context-engineering/ng-context-packgit clone --depth 1 https://github.com/FlyFission/nuclear-grade-context-engineeringWrote 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.
[](https://agentmods.dev/commands/flyfission/nuclear-grade-context-engineering/ng-context-pack)<a href="https://agentmods.dev/commands/flyfission/nuclear-grade-context-engineering/ng-context-pack"><img src="https://agentmods.dev/badge/commands/flyfission/nuclear-grade-context-engineering/ng-context-pack.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00000 | $0.00834 |
| Opus 5 | $0.00000 | $0.00417 |
| Sonnet 5 | $0.00000 | $0.00167 |
| Haiku 4.5 | $0.00000 | $0.00083 |
Grade A, and why
ng-context-pack 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.
How it starts
The opening of the file, as written. The whole thing — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ng-context-pack
Portable command prompt generated from skills/briefing-an-agent/SKILL.md. Edit the skill, then run python tools/ng.py gen-commands; do not edit this file by hand.
Prepares focused context for an AI agent, reviewer, verifier, or releaser at the start of a task, with a clear role, goal anchor, authority, evidence to produce, forbidden actions, and stop conditions. Use when an agent or reviewer is about to begin work and needs bounded context before it starts. Do not use for a tiny self-contained task that needs no briefing, or for transferring already-open work to a new owner, which is handing-off-work.
Use when
- An AI agent is about to edit files, run commands, call tools, or prepare release evidence, and needs its role, authority, and stop conditions stated before it starts.
- A reviewer needs a one-screen summary of a Standard change record before beginning the review.
- A long research or build thread has to be boiled down into the context needed to act.
Do not use when
- The task is a small Quick change, and all the context is already in
risk.mdandproof.md. - The agent has no power to act and only needs a file explained.
- Responsibility for work that is already open is transferring to a new owner. That is a handoff, not a start-of-task briefing -- use
handing-off-work, which also forces the incoming owner to restate scope, authority, and stop conditions before acting.
Inputs
- The path to the change record and the chosen mode.
- The role: builder, reviewer, verifier, releaser, incident lead, or researcher.
- The files affected, the allowed commands, the forbidden actions, the approval gates, and the evidence required.
docs/02-operating-system/context-packs.md.
Prompt text
Build a Nuclear-grade context pack for this work.
Inputs:
- packet: .nuclear/changes/<slug>/
- role: <builder|reviewer|verifier|releaser|researcher>
- decision question: <one sentence>
- objective: <one paragraph>
- work phase: <explore|candidate|audit|accept>
- affected files: <list>
- last completed action:
- changed conditions:
- critical next action and likely error:
- allowed commands/tools: <list>
- forbidden actions: <list>
- do-not-touch targets: <list>
- approval gates: <list>
- required evidence: <commands/links/reviews>
- loaded instruction files: <AGENTS.md / CLAUDE.md / .github/copilot-instructions.md / tool rules in force, with precedence or "none known">
Return a short context pack. Include the mode, the decision question, the goal, the work phase, the archetype the brief places the agent in and its characteristic drift, a risk summary, a basis summary, the evidence required, the loaded instruction files in force and which one wins on conflict, the limits on what the agent may do, the claims it must not make, the open gaps, the last action completed, what has changed, the critical next action, and the next action. If responsibility is changing hands rather than starting fresh, use the `handing-off-work` prompt instead.
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 · 60 lines · 0 tokens per session scan A 61271ae966de
ng-context-pack is a command published in the GitHub repository FlyFission/nuclear-grade-context-engineering (33 stars, last pushed 25d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 834 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.
Other commands, from other repositories
check-gates
Run a Pre-Action Gate check against prevention rules before executing a risky action.
task-verify
PitWay: Run an inprogress task's approved verification command and persist evidence.
milestone-list
PitWay: List every milestone with a concise one-line status.
fix
Universal debugging and fix application with semantic code analysis.
domain-research
Conduct domain research covering industry analysis, regulations, technology trends, and ecosystem dynamics using current web data and verified sources.
validate-prd
Validate an existing PRD against BMAD standards - comprehensive review for completeness, clarity, and quality.