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 rules/aiagentwithdhruv/ai-dev-stack/50-rag-systemgit clone --depth 1 https://github.com/aiagentwithdhruv/ai-dev-stackWhat 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 | $0.00274 |
| Opus 5 | $0.00000 | $0.00137 |
| Sonnet 5 | $0.00000 | $0.00055 |
| Haiku 4.5 | $0.00000 | $0.00027 |
Grade A, and why
50-rag-system 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 2d 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.
What it actually says
RAG architecture:
- Separate ingestion, parsing, chunking, embedding, indexing, retrieval, reranking, context assembly, and answer generation.
- Keep retrieval logic independent from answer generation logic.
- Preserve source metadata for traceability.
RAG rules:
- Never dump raw full documents into prompts when chunking is expected.
- Use deterministic chunking strategies unless explicitly experimenting.
- Maintain chunk metadata such as source, page, section, title, tenant, and timestamps where relevant.
- Prefer source attribution/citations in outputs when the product requires trust and traceability.
- Add configurable retrieval parameters such as top_k, filters, score thresholds, and reranking options.
Ingestion expectations:
- Support clean document ingestion pipelines.
- Normalize and sanitize extracted text.
- Preserve source identity and version if relevant.
- Re-embed only when necessary.
Answering expectations:
- Ground answers in retrieved context.
- Handle no-context and low-confidence cases gracefully.
- Avoid hallucinating unavailable facts.
- Return structured outputs where the product requires it.
Do not:
- Mix ingestion code with runtime answer generation in the same module.
- Make retrieval behavior impossible to inspect or tune.
- Hide retrieval failures silently.
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.
- 2d ago First seen · 34 lines · 0 tokens per session scan A 8cc5a45c07f5
50-rag-system is a cursor rule published in the GitHub repository aiagentwithdhruv/ai-dev-stack (10 stars, last pushed 2mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 274 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 2026-08-31.
Other cursor rules, from other repositories
cursorrules
You are building an AI/ML project with Python. The project uses PyTorch for model training, handles data pipelines with proper validation, tracks experiments systematically, and follows production ML engineering practices. Code is type-hinted, tested, and reproducible.
cursorrules
Code exhibits taste — Torvalds' term for finding the formulation where special cases disappear. Seek representations where edge cases become impossible, the type system proves correctness, and the API guides users into the pit of success.
java
Modern Java: records, sealed classes, streams, virtual threads.
javascript
Modern JavaScript: ES2023+, async patterns, common traps.
accessibility
Accessibility: semantic HTML, ARIA, keyboard navigation, testing.
cross-tool-config
Cross-tool AI config: what transfers between Cursor, Claude Code, Copilot, Windsurf, Gemini, and Codex.