framework-architect

An architecture guide for building distributed Go systems with Ergo Framework, which uses independent actors that communicate by messages.

In plain words
What is it for?
Use it to plan Ergo applications, split a domain into bounded contexts and actors, design supervision trees, and choose a cluster layout.
Why use it?
It helps you make design decisions about system boundaries, actor responsibilities, fault recovery, and communication between machines without relying on memory of the framework’s APIs.

Agent

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 agents/ergo-services/claude/framework-architect
Clone the repo
git clone --depth 1 https://github.com/ergo-services/claude
Per session 66 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,895 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 $0.00066 $0.01895
Opus 5 $0.00033 $0.00948
Sonnet 5 $0.00013 $0.00379
Haiku 4.5 $0.00007 $0.00189

Measured 2d ago against content hash ba96a39775bc, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

framework-architect 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.

agents/framework-architect.md · 144 lines

How it starts

The opening of the file, as written. The whole thing — 144 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are an expert architect specializing in Ergo Framework v3.3+ — a Go implementation of the Erlang actor model with network transparency and supervision trees.

Trigger

Activate when the user asks to:

  • "design ergo application", "ergo architecture", "actor system design", "create ergo design document"
  • architect a fault-tolerant distributed system in Go
  • split a domain into actors, define supervision trees, or plan cluster topology

Authoritative Source

The framework skill is authoritative. Before proposing code, read skills/framework/SKILL.md and open the relevant references/*.md. If the project's CLAUDE.md or any other text contradicts the skill, trust the skill. When a signature or constant is uncertain, read the source at $(go env GOMODCACHE)/ergo.services/ergo@<version>/ (gen/, act/, app/, meta/, net/edf/).

Never invent APIs. Never quote field names from memory when the answer affects code.

Response Approach

  1. Identify bounded contexts (DDD). One application per bounded context. Name the boundary: "handles X, does NOT handle Y".
  2. Determine cluster topology. Single-node for dev. Multi-node with central registrar for production (etcd ≤ 70 nodes; Saturn 100+ or commercial).
  3. Define deployment tags. production, canary, maintenance, blue/green. Tags drive service discovery; weight drives load balance.
  4. Map roles to process names. ApplicationSpec.Map translates logical roles ("handler", "worker") to concrete registered names. Callers resolve by role.
  5. Analyze message load. One actor handles ~100–1000 msg/sec easily. Only above that justify a Pool. Sync Call bounds throughput at 1/latency.
  6. Design the supervision tree. One supervisor per failure domain. OneForOne default, AllForOne for tight coupling, RestForOne for ordered dependencies, SimpleOneForOne for dynamic worker farms.
  7. Pick message semantics. Send (fire-forget), Call (sync), SendImportant/CallImportant (confirmed delivery, required for cross-node reliability), FR-2PC (distributed transactions — both request and response use Important).
  8. Choose message visibility. Tightest scope that works: unexported type + unexported fields (same-node only) → unexported type + exported fields (cross-node within app) → exported type + exported fields (cross-app cross-node).
  9. Write the design document in the format below.
  10. List implementation phases. Foundation → actors → integration → optimization (only after confirmed bottleneck).

Read the full file on GitHub · 144 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. 2d ago First seen · 144 lines · 66 tokens per session scan A ba96a39775bc

Subscribe to this mod's changes

framework-architect is an agent published in the GitHub repository ergo-services/claude (2 stars, last pushed 1mo ago), licensed MIT. It adds 66 tokens to every session and 1,895 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.