agentmods.dev

asya AGENTS.md

Project-specific guidance for AI coding agents working on Asya, a system that runs AI workloads on Kubernetes through message-passing actors. It covers the project structure, testing, and how messages are routed between components.

In plain words
What is it for?
Understanding Asya’s components, finding the correct files to edit, following its testing strategy, and handling message envelopes and deployment-related code.
Why use it?
It gives an agent the context and rules needed to change the project safely, reducing confusion about its architecture and source files.

Instructions file for CodexOpenCode

59 6d ago C scan 4,405 tokens original Apache-2.0
View the source file deliveryhero/asya
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 instructions/deliveryhero/asya/agents-md
Clone the repo
git clone --depth 1 https://github.com/deliveryhero/asya

Made for: Codex, OpenCode.

Per session 4,405 This file is loaded in full into every session.
When invoked 4,405 The same file — it is already loaded in full.
Security scan C 2 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 $0.04405 $0.04405
Opus 5 $0.02202 $0.02202
Sonnet 5 $0.00881 $0.00881
Haiku 4.5 $0.00441 $0.00441

Measured 3d ago against content hash eddd9b21a8d8, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

asya AGENTS.md scanned grade C with 2 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.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

**Prerequisites**: uv (`curl -LsSf https://astral.sh/uv/install.sh | sh`), Go 1.24+, Python 3.13+, Docker, Make

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

**Prerequisites**: uv (`curl -LsSf https://astral.sh/uv/install.sh | sh`), Go 1.24+, Python 3.13+, Docker, Make
AGENTS.md · 362 lines

How it starts

The opening of the file, as written. The whole thing — 362 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

AI developer guidance for the Asya project.

Project Overview

Asya is an Actor Mesh framework for running AI workloads on Kubernetes using choreography (decentralized) instead of centralized orchestration. Actors communicate by passing envelopes through message queues; routing is embedded in each envelope, not managed by a central coordinator. Implemented with Crossplane Compositions for declarative actor deployment with inline sidecar rendering.

Core components (all in src/):

  • asya-sidecar (Go): envelope router injected into actor pods; Queue → Sidecar → Runtime → Sidecar → Next Queue
  • asya-runtime (Python): lightweight socket server loaded via ConfigMap; executes user handler, returns result. Source of truth: src/asya-runtime/asya_runtime.py (single file, no deps). deploy/helm-charts/asya-crossplane/files/asya_runtime.py is a symlink — editing the source automatically reflects in the Crossplane chart's ConfigMap. No manual sync needed.
  • asya-gateway (Go): optional MCP/HTTP gateway; exposes async actor pipelines as synchronous HTTP
  • asya-crew (Python): system actors — x-sink (persist results), x-sump (DLQ handling), x-pause (checkpoint envelope to S3 and signal paused), x-resume (restore envelope from S3 and re-inject into the mesh)
  • asya-lab (Python): CLI tools (asya mcp ..., asya flow ...) for debugging and flow compilation
  • asya-testing (Python): shared test fixtures and utilities
  • asya-state-proxy (Go): optional sidecar that gives actors virtual persistent state via filesystem emulation; actors read/write /state/... paths, runtime intercepts Python file I/O and forwards to the proxy over Unix socket; proxy translates to actual storage backend (S3, GCS, Redis, NATS KV) with configurable LWW or CAS guarantees; actors remain stateless Deployments — no StatefulSets

Project structure:

asya/
├── src/
│   ├── asya-gateway/
│   ├── asya-sidecar/
│   ├── asya-runtime/
│   ├── asya-crew/
│   ├── asya-testing/
│   └── asya-cli/
├── deploy/helm-charts/
│   ├── asya-crew/          # pre-built generic actors (like x-sink, x-sump)
│   ├── asya-crossplane/    # contains AsyncActor XRD
│   └── asya-gateway/       # sync stateful HTTP gateway exposing actors and flows as MCP tools or A2A agents
├── testing/
│   ├── component/          # docker-compose
│   ├── integration/        # docker-compose
│   ├── e2e/                # local kind cluster
│   └── shared/             # shared docker-compose configurations
└── examples/
│   ├── asyas/              # sample XR definitions
    └── flows/              # sample Python flows

Read the full file on GitHub · 362 lines

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. 3d ago First seen · 362 lines · 4,405 tokens per session scan C eddd9b21a8d8

Subscribe to this mod's changes

asya AGENTS.md is an instructions file published in the GitHub repository deliveryhero/asya (59 stars, last pushed 6d ago), licensed Apache-2.0. It adds 4,405 tokens to every session, about $0.0220 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

actors-frameworkaiai-orchestrationevent-drivenkubernetes
Related

Other instructions, from other repositories

superplane AGENTS.md

AGENTS.md instructions for superplanehq/superplane, covering repository guidelines, project overview, repository layout, prerequisites & setup and build, test & lint commands.

superplanehq/superplane · 3,898 tokens

superplane CLAUDE.md

Claude Code instructions for superplanehq/superplane, a project described as: Open source factory for one-shot engineering.

superplanehq/superplane · 10 tokens

netdata AGENTS.md

AGENTS.md instructions for netdata/netdata, covering agents.md, goals, requirement language, mandatory development principles and sow system.

netdata/netdata · 14,018 tokens

agent-sandbox AGENTS.md

Instructions for kubernetes-sigs/agent-sandbox, covering agents.md, project summary, repository layout, agent skills and build, test, lint.

kubernetes-sigs/agent-sandbox · 4,218 tokens

agent-sandbox copilot-instructions.md

Instructions for kubernetes-sigs/agent-sandbox: Project Context & Architecture: Refer to AGENTS.md for full project background, module layout, toolchain versions, and core conventions.

kubernetes-sigs/agent-sandbox · 293 tokens

sortie copilot-instructions.md

Instructions for sortie-ai/sortie, covering sortie coding & review standards, 1. layered imports (downward only; violation is critical), 2. concurrency safety, 3. workspace path safety (critical: security boundary) and 4. persistence (sqlite).

sortie-ai/sortie · 1,030 tokens