Most autonomous
Most agent tools are interactive - useful for hands-on work, less so for the long tail of recurring tasks: priority briefs, market monitoring, PR reviews, research digests, security scans. aeon owns that tail and runs it unattended.
The most autonomous agent framework - no approval loops, no babysitting.
SKILLS★ZERO BABYSITTING
autoresearch, create-skill, skill-health, skill-repair, skill-evals, self-improve, spawn-instance, fleet-control, deploy-prototype, vuln-scanner, distribute-tokens, install-skill, and more.
auto-merge, pr-review, pr-triage, issue-triage, code-health, repo-scanner, workflow-audit, github-monitor, github-trending, ecosystem-pulse, fork-fleet, and more.
heartbeat, capabilities-map, skill-scan, cost-report, api-health, batch-health, operator-scorecard, fork-events, fork-health, skill-gap, product-pulse, and more.
onchain-monitor, defi-overview, monitor-polymarket, token-movers, token-pick, unlock-monitor, narrative-tracker, picks-tracker, fear-divergence, x402-monitor, base-mcp, and more.
deep-research, paper-digest, paper-pick, hn-digest, reddit-digest, telegram-digest, article, article-queue, framework-watch, narrative-convergence, fetch-tweets, and more.
action-converter, deal-flow, goal-tracker, reflect, followup-patrol, shiplog, strategy-builder, tool-builder, idea-pipeline, note-taking, send-email, and more.
reply-maker, mention-radar, farcaster-digest, write-tweet, syndicate-article, product-hunt, schedule-ads, soul-builder, and more.
investigation-report, tx-explain, vigil, wallet-profile, and more.
Other agents put you in the driver's seat - approve this tool call, review this diff, confirm this action. aeon is a system you configure and walk away from. It decides when to run, what to check, and when to bother you.
Most agent tools are interactive - useful for hands-on work, less so for the long tail of recurring tasks: priority briefs, market monitoring, PR reviews, research digests, security scans. aeon owns that tail and runs it unattended.
Every skill output is scored 1–5 by Haiku after each run. Flags like api_error, stale_data, and rate_limited live in a rolling 30-run history. heartbeat audits the fleet 3× daily; if a skill fails 3× in a row, skill-repair diagnoses and patches it.
A Next.js dashboard plus a set of GitHub Actions workflows. Fork the repo, run ./aeon, authenticate, pick skills, push. Actions handles scheduling, execution, and notifications via Telegram, Discord, Slack, or Email. Public repo = unlimited free minutes.
Two files define your agent. SOUL.md carries your voice - dump your tweets, essays, and posts in a folder and the builder distills your worldview into markdown any LLM can embody. STRATEGY.md carries your intent - one metric, a few priorities, the hard limits. Together they're the north star every one of the 102 skills reads before it acts.
Every skill reads your voice and your intent, and lets them break ties - what to prioritise, what to skip, when to bother you.
Chasing a metric it cannot measure yet? aeon writes the missing skill and asks you for the API keys it needs to pull the data.
Run after run it reprioritises the fleet and tunes its own skills - compounding toward your number with no new instruction.
On every tick aeon reads your voice, your north-star, and everything it has learned, then acts unsupervised. And it doesn't just run - it evolves: tuning its own skills, forking itself into specialised repos, and carrying the API keys across. Set the direction once and it compounds toward your goal - one agent becoming a whole self-running fleet.
$AEON is a fair launched token - no insider allocation & no unlocks. Fees flow to the treasury and back into buybacks.