60-agents

A set of design rules for building AI agents and coordinating the tools they use.

In plain words
What is it for?
Designing agent architecture, defining tool inputs and outputs, validating results, managing memory, and coordinating multiple agents.
Why use it?
It helps keep agent systems understandable, controlled, and safe as they grow. It also prevents hidden tool failures and unrestricted actions.

Cursor rule

Install

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.

agentmods
npx agentmods add rules/aiagentwithdhruv/ai-dev-stack/60-agents
Clone the repo
git clone --depth 1 https://github.com/aiagentwithdhruv/ai-dev-stack
Per session 0 Nothing until a file matches its globs; then the whole rule loads.
When invoked 260 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00260
Opus 5 $0.00000 $0.00130
Sonnet 5 $0.00000 $0.00052
Haiku 4.5 $0.00000 $0.00026

Measured 2d ago against content hash 55f775ff0042, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

60-agents 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.

rules/60-agents.mdc · 37 lines

What it actually says

Agent system architecture:

  • Separate planner, executor, tools, memory, state, and evaluation logic.
  • Agents must be modular and role-specific where possible.
  • Tool calls should be explicit, validated, and logged.

Agent rules:

  • Agents must not directly access databases or external APIs unless explicitly designed through a tool layer.
  • Prompts must be templated and stored separately from orchestration logic.
  • Keep critical business workflows deterministic where possible.
  • Add output validation and fallback behavior.
  • Use structured schemas for agent outputs in production paths.
  • Prefer supervisor/policy logic for multi-agent coordination.

Tooling rules:

  • Every tool should have:
    • a clear purpose
    • input schema
    • output schema
    • failure behavior
  • Tools should be side-effect aware.
  • Sensitive tools must enforce auth and authorization checks.

Memory/state rules:

  • Separate short-term conversational state from long-term memory.
  • Store long-term memory only when product requirements justify it.
  • Track provenance for stored memories if relevant.

Do not:

  • Let agents perform unrestricted actions.
  • Hide tool errors.
  • Mix prompt text deeply into service/business code.
Changes

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.

  1. 2d ago First seen · 37 lines · 0 tokens per session scan A 55f775ff0042

Subscribe to this mod's changes

60-agents 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 260 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.