Add Cycles to awesome-agents.yaml (Safety Guardrails)

[amavashev]

What

Adds Cycles to awesome-agents.yaml under the categories AI Agents, Safety Guardrails (Safeguarding), and Standardization.

Entry

- project: "Cycles"
  project_description: "Open-source runtime authority for AI agents — enforces hard limits on spend, risk, and tool actions before execution. Multi-tenant, concurrency-safe; SDKs for Python/TypeScript/Rust, Spring Boot starter for Java, and an MCP server compatible with Claude Desktop, Cursor, and Windsurf"
  project_is_open_source: true
  sources:
    - source: "code"
      source_url: "https://github.com/runcycles/cycles-server"
    - source: "website"
      source_url: "https://runcycles.io"
    - source: "protocol"
      source_url: "https://github.com/runcycles/cycles-protocol"
  categories:
    - "AI Agents"
    - "Safety Guardrails (Safeguarding)"
    - "Standardization"

What is Cycles

Cycles is an open-source (Apache 2.0) runtime layer that enforces hard limits on AI agent spend, risk, and tool actions before they execute. Multi-tenant by default, concurrency-safe across thousands of agents.

The reservation lifecycle (reserve / execute / commit / release) wraps any LLM or tool call and ensures agents cannot authorize more spend or take riskier actions than policy allows. The protocol is openly specified — any implementer can build a conformant server.

Categories

• Safety Guardrails (Safeguarding) is the primary fit — Cycles is a runtime enforcement / guardrail layer.
• Standardization because Cycles ships an open protocol spec (cycles-protocol) under Apache 2.0 with a conformance suite.
• AI Agents is the umbrella category.

Happy to drop or remap any of these if you'd prefer a tighter scope.

Thanks for maintaining this list!