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 instructions/secundant/neodx/agents-mdgit clone --depth 1 https://github.com/secundant/neodxWhat 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 | $0.02722 | $0.02722 |
| Opus 5 | $0.01361 | $0.01361 |
| Sonnet 5 | $0.00544 | $0.00544 |
| Haiku 4.5 | $0.00272 | $0.00272 |
Grade A, and why
neodx 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 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.
How it starts
The opening of the file, as written. The whole thing — 166 lines — stays where its author put it; the contents beside it link to each section on GitHub.
neodx
Yarn 4 / Vite+ monorepo of everyday frontend DX instruments: SVG sprite pipeline, Figma integration, isomorphic logger, and a virtual file system, plus shared foundations and build tooling.
This file is the routing index: it keeps repo-wide constraints visible and points to the narrowest source that owns the current decision.
Session rules
These bind every session until the owner changes them.
One branch. Do the work on the standing branch work (cut over 2026-08-24 from main after
strip-source-bridges landed). Do not open PRs, merge to main, or add extra branches unless the
owner asks. main is landed history. Release automation on main is not everyday workflow.
Changelogs, not publishes. When a caller-visible change needs a changelog, add a Changeset with the CLI and stop there:
yarn changeset add
Docs: Changesets CLI. This repo pins @changesets/cli 2.27.1,
so add takes --empty and --open. It does not take the v3 --patch / --minor / -m flags.
Publish freeze. strip-source-bridges (#180) is
on npm 1.1.1 and closed. Do not publish again until the owner names a consolidated release.
Queue later leftovers as Changesets on work. Do not run yarn changeset version,
yarn changeset publish, merge a Version Packages PR, recreate a closed Version Packages PR, or
re-run Release to push npm. The freeze exists so later work can share one release.
Session close. End every session with the session-close gate. Name any skipped check and why. Local green is not optional at close, even when CI is otherwise opt-in.
Task names. In issues, ledgers, and chat, lead with a readable slug, then the GitHub number:
oxlint-typecheck (#179), not R2-f. Closed
program rows may keep old stream codes as history. Live work must not.
Tighten what you touch. Config, tools, and comments have gone vague in places (stale program codes, duplicated tsconfig, dishonest overrides, dead paths, names that no longer match meaning). Every session, after the assigned slice is green, make a bounded pass over files and tools this session already opened: clarify ownership, delete dead config, collapse duplicates, and make names match meaning. Do not start a repo-wide cleanup. Do not mix this pass with pack-contract or publish work. Record leftover debt with a slug, not a silent skip. This rule outlives any one leftover.
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.
- 3d ago First seen · 166 lines · 2,722 tokens per session scan A 4030b02c208e
neodx AGENTS.md is an instructions file published in the GitHub repository secundant/neodx (94 stars, last pushed 8d ago), licensed MIT. It adds 2,722 tokens to every session, about $0.0136 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.
Other instructions, from other repositories
heroui CLAUDE.md
Claude Code instructions for heroui-inc/heroui, covering claude.md, repository overview, key technical stack, development commands and core development commands.
heroui AGENTS.md
AGENTS.md instructions for heroui-inc/heroui, covering agents.md, repository overview, tech stack, monorepo structure and commands.
react CLAUDE.md
Claude Code instructions for react/react, covering react, monorepo overview and current active work.
joint CLAUDE.md
Instructions for clientIO/joint, covering claude.md, project overview, common commands, install dependencies and build distribution files (runs grunt).
tsdown AGENTS.md
Instructions for rolldown/tsdown, covering agents.md, project overview, development commands, building and alias of pnpm build.
thesvg AGENTS.md
AGENTS.md instructions for glincker/thesvg, covering thesvg - ai review guidelines, svg icon submission rules, required, fill and color rules and icons.json registry.