slancha-wire 0.5.13

Magic-wormhole for AI agents — bilateral signed-message bus over a mailbox relay
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74

# Fly.io app config for wire-relay-server.
#
# Deploy:
#   fly launch --no-deploy           # first time: creates the app
#   fly volumes create relay_state \
#       --size 1 --region iad        # 1 GB persistent volume; wire-state lives here
#   fly deploy                       # build + push + roll
#   fly status                       # health
#
# Cutover:
#   In Cloudflare DNS for wireup.net:
#     - CNAME relay → <app>.fly.dev  (proxied / orange-cloud for DDoS + caching)
#     - CNAME @ (apex) → <app>.fly.dev  OR keep apex on CF Pages for landing
#   Then point new clients at `https://relay.wireup.net` and let DNS
#   propagation drain old wireup.net traffic from Spark cloudflared tunnel.
#
# Pre-launch sanity:
#   curl https://<app>.fly.dev/healthz                                   # → ok
#   curl https://<app>.fly.dev/.well-known/agent-card.json?handle=test   # → 404 fine
#
# Cost @ v0.5 scale: $0 on Fly hobby plan (shared-cpu-1x 256MB + 1 GB volume covered).
# Bump memory / region count when you outgrow it.

app = "wireup-relay"
primary_region = "iad"   # change if your CF edge sits elsewhere (sjc / fra / etc.)

[build]
# Empty — Fly uses the repo-root Dockerfile.

[env]
# wire-relay-server reads WIRE_HOME for slot state. Already set in the
# Dockerfile, but re-stating here keeps Fly's config the single source
# of truth for environment.
WIRE_HOME = "/data"
RUST_LOG = "wire=info"

[http_service]
internal_port = 8770
force_https = true

# SSE / streaming: never auto-stop machines while subscribers are
# connected. min_machines_running = 1 keeps the relay warm.
auto_stop_machines = false
auto_start_machines = true
min_machines_running = 1

# Long timeouts so SSE keepalives don't disconnect.
# (Fly's default 60s idle timeout would kill our 30s-keepalive streams.)
[http_service.concurrency]
type = "requests"
soft_limit = 200
hard_limit = 250

[[mounts]]
source = "relay_state"
destination = "/data"
# initial_size only applies on first attach; resize later with `fly volumes extend`.
initial_size = "1gb"

[[vm]]
size = "shared-cpu-1x"
memory = "256mb"
cpus = 1
# Bump to "shared-cpu-2x" + 512mb when SSE subscriber count >100.

[checks]
  [checks.healthz]
  type = "http"
  port = 8770
  path = "/healthz"
  interval = "15s"
  timeout = "5s"
  grace_period = "10s"
  method = "GET"