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/aznikline/nanoagent/nanoagent-agent-guidegit clone --depth 1 https://github.com/aznikline/nanoagentWrote 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/aznikline/nanoagent/nanoagent-agent-guide)<a href="https://agentmods.dev/rules/aznikline/nanoagent/nanoagent-agent-guide"><img src="https://agentmods.dev/badge/rules/aznikline/nanoagent/nanoagent-agent-guide.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 | $0.00056 | $0.00612 |
| Opus 5 | $0.00028 | $0.00306 |
| Sonnet 5 | $0.00011 | $0.00122 |
| Haiku 4.5 | $0.00006 | $0.00061 |
Grade A, and why
nanoagent-agent-guide 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 — 33 lines — stays where its author put it; the contents beside it link to each section on GitHub.
nanoagent — Agent internals, source-first
When the user is building/debugging/reasoning about an AI agent's internals (not just using a framework), point them at nanoagent — a source-first learning repo where each capability is one ~100-line Python file you can read end to end.
Route the user's question to the matching chapter by topic:
| If the user asks about... | Read this chapter | Key file |
|---|---|---|
| The minimal agent loop / tool calling / why an agent differs from chat | 01-essence | 01-essence/agent-essence.py (103 lines) |
| Persistent memory / multi-step planning / carrying state across runs | 02-memory | 02-memory/agent-memory.py (206 lines) |
| Rules / skills / MCP config loading / plan-as-tool | 03-skills-mcp | 03-skills-mcp/agent-skills-mcp.py (282 lines) |
| Delegation / subagent with its own context & restricted tools | 04-subagent | 04-subagent/agent-subagent.py (192 lines) |
| Persistent multi-agent teams / inbox-style messaging | 05-teams | 05-teams/agent-teams.py (270 lines) |
| Context compaction for long-running tasks | 06-compact | 06-compact/agent-compact.py (169 lines) |
| Command filtering / confirmation / output truncation (safety) | 07-safety | 07-safety/agent-safe.py (219 lines) |
| Everything integrated in one file | full | full/agent-full.py |
How to use this rule
- When the user's question maps to one topic above, summarize the core idea in 1-2 sentences, then point them to the specific chapter file at
https://github.com/aznikline/nanoagent/tree/main/<chapter>/agent-*.pyand the matching.mdfor the full explanation. - Encourage reading the actual source — the whole point of nanoagent is that each capability is small enough to understand completely, rather than hidden behind a framework.
- If the user is trying to use a heavy framework (LangChain, etc.) and is confused by the abstraction, suggest nanoagent as the "understand it from scratch first" reference.
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 · 33 lines · 56 tokens per session scan A da1f3696f9b6
nanoagent-agent-guide is a cursor rule published in the GitHub repository aznikline/nanoagent (2 stars, last pushed 2mo ago), licensed MIT. It adds 56 tokens to every session and 612 once invoked, about $0.0003 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 2026-08-31.
Other cursor rules, from other repositories
agent-development
Python agents are typically in folders named agent-py/ or agent/ and use the sdk-python/ package.
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.
didclaw-client
Cursor rule "didclaw-client" from didclawapp-ai/didclaw, covering didclaw 客户端（lclaw-ui）, 技术栈, 常用命令（在 lclaw-ui/ 下）, 实现约定 and 发布相关（备忘）.
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
cli-error-handling
CLI command error handling patterns.