[management,proxy] Agent network: per-account LLM gateway (policy, metering, multi-provider) (#6555)

* [agent-network] Shared proto, OpenAPI schema, and generated types

* [agent-network] Management: store, manager, synthesizer, policy engine, provider catalog, HTTP/gRPC API

Adds the account-scoped agent-network module: provider/policy/budget CRUD and
store, the reverse-proxy service synthesizer, policy selection + limit
enforcement, the provider catalog (incl. Vertex AI and AWS Bedrock entries),
and the management HTTP + proxy gRPC surfaces.

* [management] Fix agent-network proxy-peer fan-out on affected-peer recompute

The affected-peers resolver loaded only persisted reverse-proxy services, but
agent-network services are synthesized on demand and never persisted. As a
result the embedded proxy peer was never folded into the affected set when a
client's group changed, so the proxy received no network-map update for a newly
authorised client and rejected its handshake until a full resync (restart).

loadProxyServices now merges the synthesized agent-network services (injected
via a registration hook to avoid an import cycle), so proxy peers learn newly
authorised clients immediately.

* [proxy] Reverse-proxy middleware framework, chain, and request plumbing

The per-target middleware chain (slots, dispatcher, mutation gate, metadata
merger), body capture, access-log terminal sink, and the proxy wiring that
builds + runs chains for synthesized agent-network services.

* [proxy] LLM parsers, pricing, and builtin middlewares (OpenAI, Anthropic, Vertex AI, AWS Bedrock)

Request/response parsers and SSE/event-stream metering, the embedded pricing
table, and the builtin middleware set: request parser, router, policy
limit-check/record, cost meter, guardrail, identity inject, response parser.
Includes the path-routed providers — Google Vertex AI (keyfile:: service-account
OAuth minting) and AWS Bedrock (bearer auth, invoke/converse/streaming, optional
/bedrock prefix) — plus the Models allowlist and unmeterable-publisher deny.

* [proxy] IPv6 in-place apply and TCP accept-loop hardening on netstack listeners

* [agent-network] End-to-end test suite, module docs, and deployment preset

* [agent-network] Fix codespell typos and exclude false positives

- labelgen word pool: vermillion -> vermilion, racoon -> raccoon.
- codespell ignore list: add flate (Go compress/flate package), recordin
  (a test-local identifier), and unparseable (a valid alternative spelling used
  consistently across identifiers + a metadata-value constant).

* [management] Set LastSeen on injected proxy peer in realstack test (MySQL strict-mode)

The injected embedded proxy peer had a PeerStatus with a zero LastSeen, which
serializes to '0000-00-00' and is rejected by MySQL in strict mode (SQLite
tolerates it). Set LastSeen to a valid time so SaveAccount succeeds on both
engines.

* [agent-network] Remove e2e shell-script suite from this branch

The end-to-end shell scripts under scripts/e2e/ are maintained in a separate
testing suite and are not part of this change set.

* [agent-network] Polish module docs: remove internal review scaffolding, fix links, verify diagrams

Strip PR-review framing, commit references, absolute paths, and stale internal
references from the agent-network module docs; fix broken relative links; verify
all diagrams against the current architecture. Remove the internal AI-reviewer
prompt file.

* [management] Refine session expiration handling to support 3-state encoding for SSO deadlines

* [agent-network] Relocate agentnetwork package to internals/modules

Move management/server/agentnetwork (and its catalog/, labelgen/, types/
subpackages) to management/internals/modules/agentnetwork, alongside the
reverse-proxy module, and rewrite all importers. Pure relocation: package names,
the synthesizer + affectedpeers registration hook, and store access (shared
store.Store) are unchanged, so no import cycle is introduced (affectedpeers
still depends only on the agentnetwork/types leaf).

* [agent-network] Co-locate HTTP handlers in the module (RegisterEndpoints)

Move the agent-network HTTP handlers from server/http/handlers/agentnetwork into
the module at internals/modules/agentnetwork/handlers (package handlers) and
rename the entrypoint AddEndpoints -> RegisterEndpoints, matching the
reverse-proxy module convention. Wiring in http/handler.go updated accordingly.
This commit is contained in:
Maycon Santos
2026-06-27 13:41:00 +02:00
committed by GitHub
parent 615631567a
commit b416063bcc
187 changed files with 36835 additions and 660 deletions
+70 -4
View File
@@ -8,6 +8,8 @@ import (
"net"
"net/http"
"net/netip"
"os"
"strings"
"sync"
"time"
@@ -347,8 +349,20 @@ func (n *NetBird) createClientEntry(ctx context.Context, accountID types.Account
"public_key": publicKey.String(),
}).Info("proxy peer authenticated successfully with management")
// Embedded client log level: warn by default (quiet in production); set
// NB_PROXY_CLIENT_LOG_LEVEL (e.g. "trace") to surface the embedded NetBird
// client's relay / signal / handshake detail for local debugging.
clientLogLevel := log.WarnLevel.String()
if v := strings.TrimSpace(os.Getenv("NB_PROXY_CLIENT_LOG_LEVEL")); v != "" {
if lvl, err := log.ParseLevel(v); err == nil {
clientLogLevel = lvl.String()
} else {
n.logger.Warnf("invalid NB_PROXY_CLIENT_LOG_LEVEL %q, using %q: %v", v, clientLogLevel, err)
}
}
n.initLogOnce.Do(func() {
if err := util.InitLog(log.WarnLevel.String(), util.LogConsole); err != nil {
if err := util.InitLog(clientLogLevel, util.LogConsole); err != nil {
n.logger.WithField("account_id", accountID).Warnf("failed to initialize embedded client logging: %v", err)
}
})
@@ -356,11 +370,11 @@ func (n *NetBird) createClientEntry(ctx context.Context, accountID types.Account
// Create embedded NetBird client with the generated private key.
// The peer has already been created via CreateProxyPeer RPC with the public key.
wgPort := int(n.clientCfg.WGPort)
client, err := embed.New(embed.Options{
embedOpts := embed.Options{
DeviceName: deviceNamePrefix + n.proxyID,
ManagementURL: n.clientCfg.MgmtAddr,
PrivateKey: privateKey.String(),
LogLevel: log.WarnLevel.String(),
LogLevel: clientLogLevel,
BlockInbound: n.clientCfg.BlockInbound,
// The embedded proxy peer must never be a stepping stone into
// the proxy host's LAN: it only exists to reach NetBird mesh
@@ -371,7 +385,9 @@ func (n *NetBird) createClientEntry(ctx context.Context, accountID types.Account
WireguardPort: &wgPort,
PreSharedKey: n.clientCfg.PreSharedKey,
Performance: n.clientCfg.Performance,
})
}
logEmbedOptions(n.logger, accountID, serviceID, publicKey.String(), embedOpts)
client, err := embed.New(embedOpts)
if err != nil {
return nil, fmt.Errorf("create netbird client: %w", err)
}
@@ -847,3 +863,53 @@ func DirectUpstreamFromContext(ctx context.Context) bool {
v, _ := ctx.Value(directUpstreamContextKey{}).(bool)
return v
}
// logEmbedOptions emits a single structured INFO line summarising every
// operationally meaningful flag handed to embed.New for this per-account
// client. Secrets (PrivateKey, PreSharedKey) are reduced to a "present"
// boolean — never logged verbatim. Use this when an embedded peer
// silently misbehaves: most failure modes (inbound drops, wrong
// management URL, v6 unexpectedly on, userspace flipped, port clash)
// are obvious from these flags before any traffic flows.
func logEmbedOptions(logger *log.Logger, accountID types.AccountID, serviceID types.ServiceID, publicKey string, opts embed.Options) {
wgPort := 0
if opts.WireguardPort != nil {
wgPort = *opts.WireguardPort
}
mtu := uint16(0)
if opts.MTU != nil {
mtu = *opts.MTU
}
perfBuffers := uint32(0)
if opts.Performance.PreallocatedBuffersPerPool != nil {
perfBuffers = *opts.Performance.PreallocatedBuffersPerPool
}
perfBatch := uint32(0)
if opts.Performance.MaxBatchSize != nil {
perfBatch = *opts.Performance.MaxBatchSize
}
logger.WithFields(log.Fields{
"account_id": accountID,
"service_id": serviceID,
"public_key": publicKey,
"device_name": opts.DeviceName,
"management_url": opts.ManagementURL,
"log_level": opts.LogLevel,
"wg_port": wgPort,
"mtu": mtu,
"block_inbound": opts.BlockInbound,
"block_lan_access": opts.BlockLANAccess,
"disable_ipv6": opts.DisableIPv6,
"disable_client_routes": opts.DisableClientRoutes,
"no_userspace": opts.NoUserspace,
"config_path_set": opts.ConfigPath != "",
"state_path_set": opts.StatePath != "",
"private_key_present": opts.PrivateKey != "",
"presharedkey_present": opts.PreSharedKey != "",
"setup_key_present": opts.SetupKey != "",
"jwt_token_present": opts.JWTToken != "",
"dns_labels": opts.DNSLabels,
"perf_buffers_per_pool": perfBuffers,
"perf_max_batch_size": perfBatch,
}).Info("starting embedded netbird client for account")
}