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/devapro/n-dashboard/claude-mdgit clone --depth 1 https://github.com/devapro/n-dashboardWhat 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 | $0.02871 | $0.02871 |
| Opus 5 | $0.01435 | $0.01435 |
| Sonnet 5 | $0.00574 | $0.00574 |
| Haiku 4.5 | $0.00287 | $0.00287 |
Grade A, and why
n-dashboard CLAUDE.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 yesterday.
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 — 221 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Notes for anyone — human or agent — working in this repository. README.md is the
user-facing documentation; this file is the parts that are easy to break.
What this is
A personal dashboard for a trusted LAN. One idea carries the whole design: a widget is a TypeScript function on disk. The platform decides when to call it (clock, MQTT message, webhook), validates what it returns against a typed schema, and pushes the result to every open tab over SSE. There is no plugin format and no config-driven renderer path, so summing a field, joining two APIs and remembering what you already saw are all the same amount of work.
Widget code is trusted. Isolation exists to contain bugs, not attackers. There
is one optional login, over /admin only (ADMIN_PASSWORD in .env) — it keeps
the panel out of reach of whoever walks past the wall display, and it is not a
hardening story. Multi-user is still out of scope; see the bottom.
Commands
npm run dev # node --watch on :8080 + vite on :5273 (proxies /api)
npm test # vitest; MQTT suites skip themselves without a broker
npm run typecheck # both projects; run this, tsc catches what vitest can't
npm run build # web bundle only — the server has no build step
docker compose -p n-dashboard-dev -f docker/compose.dev.yml up -d # broker for tests
Repo map
.claude/skills/ write-widget — the widget-authoring workflow, for agents
src/shared/ ctx.d.ts (the widget API), kind schemas, duration parsing
src/server/ express, worker pool, scheduler, connections, ndjson store
src/web/ react — dashboard (kept light) + admin (Monaco, lazy-loaded)
docker/ Dockerfile, both compose files, broker configs
docs/ the long-form guides linked from README.md, plus the Pages page
examples/ seeded into data/ disabled on first run
data/ runtime state, gitignored
Invariants
Break one of these and it usually fails somewhere far away from the change.
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.
- yesterday First seen · 221 lines · 2,871 tokens per session scan A 94b9ebc11889
n-dashboard CLAUDE.md is an instructions file published in the GitHub repository devapro/n-dashboard (11 stars, last pushed 1mo ago), licensed MIT. It adds 2,871 tokens to every session, about $0.0144 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
dograh AGENTS.md
Instructions for dograh-hq/dograh, covering dograh - project overview, project structure, tech stack, local development and environment configuration.
grok-register-panel AGENTS.md
Instructions for lij768423-svg/grok-register-panel, covering agents.md, project scope, architecture and ownership, runtime flows and setup and run.
homelab-monitor CLAUDE.md
Instructions for SikamikanikoBG/homelab-monitor, covering claude.md and conventions.
piper CLAUDE.md
Instructions for piperbox/piper, covering claude.md, what this is, coding principles, 1. think before coding and 2. simplicity first.
oh-my-hi CLAUDE.md
Instructions for netil/oh-my-hi, covering oh-my-hi, tech stack, build & test, dev workflow and build gotcha.
piper AGENTS.md
Instructions for piperbox/piper: Read CLAUDE.md in full first — it is the authoritative source for coding principles, commands, constraints, layering, and branch/PR/issue workflow. This file only captures high-signal facts an agent would otherwise get wrong that aren't obvious from filenames or CLAUDE.md.