axon v1.25.2-axon.7

Full Changelog: v1.25.2-axon.6...v1.25.2-axon.7

axon v1.25.2-axon.6

Full Changelog: v1.25.2-axon.5...v1.25.2-axon.6

axon v1.25.2-axon.5

Full Changelog: v1.25.2-axon.2...v1.25.2-axon.5

axon v2.28.0-axon.3

Full Changelog: v2.28.0-axon.2...v2.28.0-axon.3

axon v1.25.2-axon.2

What's Changed

• build(deps): bump golang.org/x/crypto from 0.56.0 to 0.57.0 by @dependabot[bot] in #1

New Contributors

• @dependabot[bot] made their first contribution in #1

Full Changelog: v2.28.0-axon.3...v1.25.2-axon.2

axon v2.28.0-axon.2

Full Changelog: v2.28.0-axon.1...v2.28.0-axon.2

axon v2.28.0-axon.1 — first public release

First public release of axon, the AI-native fork of ntfy.

Based on upstream ntfy v2.28.0 — drop-in compatible: same wire protocol, binary name, and config paths. Unmodified ntfy apps work against it.

What's inside (all opt-in, off by default)

• AI subscription wizard — describe what you want; review and apply the proposed topics/filters/priorities
• AI tuning — 'only page me for real emergencies at night'
• Inline enrichment — summaries as titles + importance classification, instant pass-through when the AI is slow
• Digests & scheduled briefings — per-topic digests, cross-topic briefings, daily scheduled delivery in your timezone to a private topic
• Chat over history — cited, streaming answers about your recent notifications
• MCP server (ntfy mcp) — publish, wait for replies, read history, ask questions; scoped agent tokens
• Providers: Poolside (Laguna), Ollama (local-first), OpenAI-compatible, Anthropic, mock — with response cache, per-visitor + global daily token budgets, and full audit logging

Install (Linux)

curl -LO https://github.com/mptfire/axon/releases/download/v2.28.0-axon.1/axon-linux-amd64.gz
gunzip axon-linux-amd64.gz && chmod +x axon-linux-amd64 && sudo mv axon-linux-amd64 /usr/local/bin/axon
axon serve --listen-http :8080            # plain ntfy mode (AI off)

arm64 build included. Web app, docs, and the AI layer are embedded in the binary.

Enable AI

# /etc/ntfy/server.yml — local-first with Ollama:
ai-enabled: true
ai-provider: ollama
ai-base-url: http://localhost:11434/v1
ai-model: llama3.1

Or ai-provider: poolside / openai / anthropic with ai-api-key-file. See docs/config.md.

Security note: published after a full security & privacy audit (docs/security-audit.md) — no secrets or personal data in the repository or its history.