SNX1FHC OO5K9UE5How aeon is built - the operator config, identity and memory, and channels that feed a runtime spine on GitHub Actions, the eight-way LLM gateway cascade, the self-healing loop, and fleet replication.

aeon runs the perception → action → observation loop unattended on GitHub Actions -81 scheduled skills, persistent memory, a self-healing repair loop, and fleet replication.

One map, three bands

Three input layers - operator config, identity/memory, and channels and MCP servers - feed a single left-to-right runtime spine. The spine drops into the model gateway, a fleet it can spawn, and a loop that heals it.

For devs · The scheduler (scheduler.yml, */5) dispatches every due skill in aeon.yml. CLAUDE.md, STRATEGY.md,soul/ and memory/ load into context; .mcp.json servers register as tools. Output commits back to the repo and fans out via ./notify; scored runs feed the self-healing loop, and spawn-instance/fleet-control extend the fleet.

Operator Layer

Dashboard

• aeon.yml
• Skills · cron · chains · reactive triggers

Identity & Knowledge

CLAUDE.md

• Agent identity · auto-loaded every run
• Strategy + Soul
• North-star · voice · rides every run

Memory & State

• MEMORY.md · topics · logs - committed to repo by default, or issues-as-state (STATE_BACKEND); persists across runs

Channels & Interfaces

• ./notify
• Telegram · Discord · Slack · Email
• Inbound
• 5-min poll or ~1s webhook

.mcp.json

• GitHub · Base · Executor · custom MCP servers · one-click OAuth, auto-loaded into runs

Runtime - Perception · Action · Observation · unattended on GitHub Actions

Scheduler

• scheduler.yml · cron */5 · dispatches every due skill

Auth Gate

• Fleet Watcher · ALLOW / BLOCK preflight

Skill Runtime

• Claude Code (default) or any of 10 harnesses runs the matched SKILL.md prompt

Tools

• bash · gh · web

Output

• PRs · committed memory · notify · dashboard feed · Langfuse traces (opt-in)

LLM Gateway Cascade

claudeanthropicopenrouterbankrusepodvenicesurplusgrokdirect · auto-failover · Sonnet 5 default

Self-Healing Loop

Heartbeat

• daily · detects failed or chronically broken skills

Health + Scoring

• Haiku scores 1-5 · flags weak & failed runs

Repair → Self-Improve

• Health skills file votable issues (human 👍/👎 sets repair priority) · repair skills close them · prompts evolve

Fleet & Interop Layer

• spawn-instance

Forks a specialized copy · isolated billing

• fleet-control

Health-check & dispatch across instances

• MCP server

Skills as Claude Desktop tools

• fork-fleet

Tracks public forks running in the wild

Wired live

The same map, as a circuit

The same system, live - coral pulses trace how a run moves through it: in from the inputs, along the spine, out to the loops.

• configures
• context
• registers tools
• preflight
• spawns
• scores 1–5

The runtime spine

What one run does

A single run reads left to right: the scheduler picks a skill, an optional auth gate clears it, Claude Code executes it, tools do the work, and the output ships. Every model call routes through the gateway below.

• For devs · The optional Fleet Watcher preflight can ALLOW/BLOCK a run before Claude starts (no-op if secrets are unset). Claude Code executes SKILL.md with memory/ in context; tools are bash, gh, WebSearch/WebFetch, and MCP servers. Output: a PR, committed memory/ state, a ./notify fan-out, or a dashboard feed card.

The LLM gateway cascade

Every call fails over

aeon can power Claude Code eight ways. Whichever provider you rank highest and have a key for serves the run; if it dies - no credits, rate limit, outage, a dud response - the run falls through to the next, down to the direct Anthropic API.

The run loop

What happens every five minutes

Nothing runs on a server. A clock ticks; the schedule picks exactly one skill; a throwaway machine boots; Claude reads that skill's instructions, does the work, grades its own output, and commits the result back to the repo.

The self-healing loop

The fleet repairs itself

aeon watches its own work. Every run gets a grade, and the grades pile up in one file. A sentinel reads them, and when a skill keeps failing, aeon files an issue, writes a fix, opens a pull request, and merges itself back to health.

Skill chains

Skills snap together

Skills compose like building blocks: run two in parallel, drop their output into a shared folder, and hand both to a third skill that summarizes and notifies. No orchestrator, no glue code - just YAML.

It spawns its own fleet

One aeon becomes many

On a schedule, aeon can fork itself into a brand-new repo, copy the API keys across, and stand up a fleet of specialized instances, each running on its own.

Self-evolution

It grows toward your number

Point it at a metric no skill can measure yet: run create-skill, it drafts the missing skill, flags the API keys it needs, and opens a PR to wire it in. Merge, and the fleet is bigger. The scheduled self-improve pass then tunes the skills you already have, run after run.