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.
git clone --depth 1 https://github.com/lacymorrow/paperclip-hubWrote 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/rules/lacymorrow/paperclip-hub/dont-do)<a href="https://agentmods.dev/rules/lacymorrow/paperclip-hub/dont-do"><img src="https://agentmods.dev/badge/rules/lacymorrow/paperclip-hub/dont-do/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/rules/lacymorrow/paperclip-hub/dont-do"><img src="https://agentmods.dev/badge/rules/lacymorrow/paperclip-hub/dont-do.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00000 | $0.00708 |
| Opus 5.5 | $0.00000 | $0.00283 |
| Sonnet 5.5 | $0.00000 | $0.00142 |
| Haiku 4.5 | $0.00000 | $0.00071 |
Grade A, and why
dont-do 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 — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Don't Do These Things
Next.js
- Prefer use pages router (use app router)
- Don't nest server components in client components
- Don't use
use clientin server components - Don't fetch data with server actions
- Don't mix client and server code
- Don't use getStaticProps or getServerSideProps
- Don't use router.push in server components
React
- Don't use class components
- Don't use default exports
- Don't use React.FC type
- Don't use inline styles (use Tailwind)
- Don't use useState when useReducer is clearer
- Don't mutate state directly
- Don't use indexes as keys
TypeScript
- Don't use
anytype - Don't disable TypeScript checks
- Don't use
@ts-ignore - Don't use non-null assertion operator
- Don't use
Objecttype - Don't use
Functiontype - Don't use
{}as a type
State Management
- Don't use global state for local concerns
- Don't prop drill more than 2 levels
- Don't mutate context values directly
- Don't use Redux (use context + reducers)
- Don't store derived state
- Don't store server state client-side
- Don't duplicate state
API
- Don't expose internal APIs publicly
- Don't handle errors silently
- Don't return raw error messages
- Don't trust client-side data
- Don't expose sensitive information
- Don't make unnecessary API calls
- Don't ignore API errors
Security
- Don't store secrets in code
- Don't use eval() or new Function()
- Don't trust user input
- Don't expose stack traces
- Don't store sensitive data client-side
- Don't use innerHTML
- Don't disable security headers
Performance
- Don't bundle unused code
- Don't load unnecessary resources
- Don't block the main thread
- Don't ignore memory leaks
- Don't skip code splitting
- Don't ignore performance metrics
- Don't render unnecessary components
Database
- Don't use raw SQL queries
- Don't store passwords in plain text
- Don't use boolean fields (use dates)
- Don't ignore database indexes
- Don't leak connection pools
- Don't ignore transaction boundaries
- Don't store large blobs
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 · 108 lines · 708 tokens per session scan A c3660193ee13
dont-do is a cursor rule published in the GitHub repository lacymorrow/paperclip-hub (15 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 708 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 2026-11-10.
dont-do compared with cursor rules from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| dont-dolacymorrow/paperclip-hub | Cursor rule | 15 | yesterday | MIT | A | 0 | Cursor |
| update-ai-sdkgiselles-ai/giselle | Cursor rule | 560 | yesterday | Apache-2.0 | A | 11 | Cursor |
| functionsthedaviddias/llms-txt-hub | Cursor rule | 926 | yesterday | none | A | 506 | Cursor |
| technical-stackKaelSensei/MagicAIBuilder | Cursor rule | 2 | today | MIT | A | 1,259 | Cursor |
| typescript-code-generation-rulesWebDevPeterGriffin/The-VibeCode-Bible | Cursor rule | 42 | 7d ago | none | A | 402 | Cursor |
| typescript-coding-styleWebDevPeterGriffin/The-VibeCode-Bible | Cursor rule | 42 | 7d ago | none | A | 128 | Cursor |
| react-and-typescript-general-rulesWebDevPeterGriffin/The-VibeCode-Bible | Cursor rule | 42 | 7d ago | none | A | 142 | Cursor |