mirror of
https://github.com/netbirdio/netbird.git
synced 2026-07-18 20:49:56 +00:00
* [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.
170 lines
6.5 KiB
Go
170 lines
6.5 KiB
Go
package llm_response_parser
|
|
|
|
import (
|
|
"context"
|
|
"testing"
|
|
|
|
"github.com/stretchr/testify/assert"
|
|
"github.com/stretchr/testify/require"
|
|
|
|
"github.com/netbirdio/netbird/proxy/internal/middleware"
|
|
)
|
|
|
|
func TestInvoke_OpenAIStreamingWithUsage(t *testing.T) {
|
|
m := newTestMiddleware(t)
|
|
body := loadFixture(t, "openai_stream.txt")
|
|
|
|
in := &middleware.Input{
|
|
Slot: middleware.SlotOnResponse,
|
|
Status: 200,
|
|
RespHeaders: []middleware.KV{{Key: "Content-Type", Value: "text/event-stream"}},
|
|
RespBody: body,
|
|
Metadata: []middleware.KV{
|
|
{Key: middleware.KeyLLMProvider, Value: "openai"},
|
|
{Key: middleware.KeyLLMModel, Value: "gpt-4o-mini"},
|
|
},
|
|
}
|
|
|
|
out, err := m.Invoke(context.Background(), in)
|
|
require.NoError(t, err, "Invoke must not error on streaming OpenAI body")
|
|
|
|
in123, _ := metaValue(out.Metadata, middleware.KeyLLMInputTokens)
|
|
assert.Equal(t, "123", in123, "input tokens must come from final-chunk usage block")
|
|
|
|
outTok, _ := metaValue(out.Metadata, middleware.KeyLLMOutputTokens)
|
|
assert.Equal(t, "45", outTok, "output tokens must come from final-chunk usage block")
|
|
|
|
totTok, _ := metaValue(out.Metadata, middleware.KeyLLMTotalTokens)
|
|
assert.Equal(t, "168", totTok, "total tokens must come from final-chunk usage block")
|
|
|
|
completion, ok := metaValue(out.Metadata, middleware.KeyLLMResponseCompletion)
|
|
require.True(t, ok, "completion must be emitted for streaming responses")
|
|
assert.Equal(t, "Hello, world!", completion, "deltas must concatenate into the buffered fixture's text")
|
|
}
|
|
|
|
func TestInvoke_OpenAIStreamingWithoutUsage(t *testing.T) {
|
|
body := []byte(`data: {"choices":[{"delta":{"content":"Hi"}}]}
|
|
|
|
data: {"choices":[{"delta":{"content":" there"}}]}
|
|
|
|
data: [DONE]
|
|
|
|
`)
|
|
|
|
usage, completion := accumulateOpenAIStream(body)
|
|
assert.Equal(t, int64(0), usage.InputTokens, "input tokens must stay zero without a usage frame")
|
|
assert.Equal(t, int64(0), usage.OutputTokens, "output tokens must stay zero without a usage frame")
|
|
assert.Equal(t, int64(0), usage.TotalTokens, "total tokens must stay zero without a usage frame")
|
|
assert.Equal(t, "Hi there", completion, "deltas must still accumulate when usage is absent")
|
|
}
|
|
|
|
func TestInvoke_OpenAIStreamingNoUsage_OmitsUsageMetadata(t *testing.T) {
|
|
m := newTestMiddleware(t)
|
|
body := []byte(`data: {"choices":[{"delta":{"content":"Hello"}}]}
|
|
|
|
data: [DONE]
|
|
|
|
`)
|
|
in := &middleware.Input{
|
|
Slot: middleware.SlotOnResponse,
|
|
Status: 200,
|
|
RespHeaders: []middleware.KV{{Key: "Content-Type", Value: "text/event-stream"}},
|
|
RespBody: body,
|
|
Metadata: []middleware.KV{{Key: middleware.KeyLLMProvider, Value: "openai"}},
|
|
}
|
|
out, err := m.Invoke(context.Background(), in)
|
|
require.NoError(t, err, "Invoke must not error on usage-less streams")
|
|
|
|
_, hasIn := metaValue(out.Metadata, middleware.KeyLLMInputTokens)
|
|
_, hasOut := metaValue(out.Metadata, middleware.KeyLLMOutputTokens)
|
|
_, hasTot := metaValue(out.Metadata, middleware.KeyLLMTotalTokens)
|
|
assert.False(t, hasIn, "input tokens omitted when no usage frame")
|
|
assert.False(t, hasOut, "output tokens omitted when no usage frame")
|
|
assert.False(t, hasTot, "total tokens omitted when no usage frame")
|
|
|
|
completion, ok := metaValue(out.Metadata, middleware.KeyLLMResponseCompletion)
|
|
require.True(t, ok, "completion must still be emitted from deltas")
|
|
assert.Equal(t, "Hello", completion, "completion must come from delta accumulation")
|
|
}
|
|
|
|
func TestInvoke_AnthropicStreaming(t *testing.T) {
|
|
m := newTestMiddleware(t)
|
|
body := loadFixture(t, "anthropic_stream.txt")
|
|
|
|
in := &middleware.Input{
|
|
Slot: middleware.SlotOnResponse,
|
|
Status: 200,
|
|
RespHeaders: []middleware.KV{{Key: "Content-Type", Value: "text/event-stream"}},
|
|
RespBody: body,
|
|
Metadata: []middleware.KV{
|
|
{Key: middleware.KeyLLMProvider, Value: "anthropic"},
|
|
{Key: middleware.KeyLLMModel, Value: "claude-sonnet-4-5"},
|
|
},
|
|
}
|
|
|
|
out, err := m.Invoke(context.Background(), in)
|
|
require.NoError(t, err, "Invoke must not error on streaming Anthropic body")
|
|
|
|
in123, _ := metaValue(out.Metadata, middleware.KeyLLMInputTokens)
|
|
assert.Equal(t, "123", in123, "input tokens must come from message_start usage")
|
|
|
|
outTok, _ := metaValue(out.Metadata, middleware.KeyLLMOutputTokens)
|
|
assert.Equal(t, "45", outTok, "output tokens must come from message_delta usage")
|
|
|
|
totTok, _ := metaValue(out.Metadata, middleware.KeyLLMTotalTokens)
|
|
assert.Equal(t, "168", totTok, "total tokens must be input+output for anthropic streaming")
|
|
|
|
completion, ok := metaValue(out.Metadata, middleware.KeyLLMResponseCompletion)
|
|
require.True(t, ok, "completion must be emitted from text_delta accumulation")
|
|
assert.Equal(t, "Hello, world!", completion, "anthropic streaming text must accumulate across content_block_delta events")
|
|
}
|
|
|
|
func TestInvoke_StreamingTruncatedBody_BestEffort(t *testing.T) {
|
|
m := newTestMiddleware(t)
|
|
full := loadFixture(t, "anthropic_stream.txt")
|
|
cut := len(full) / 2
|
|
truncated := full[:cut]
|
|
|
|
in := &middleware.Input{
|
|
Slot: middleware.SlotOnResponse,
|
|
Status: 200,
|
|
RespHeaders: []middleware.KV{{Key: "Content-Type", Value: "text/event-stream"}},
|
|
RespBody: truncated,
|
|
RespBodyTruncated: true,
|
|
Metadata: []middleware.KV{{Key: middleware.KeyLLMProvider, Value: "anthropic"}},
|
|
}
|
|
|
|
require.NotPanics(t, func() {
|
|
_, err := m.Invoke(context.Background(), in)
|
|
require.NoError(t, err, "truncated streaming body must not surface as error")
|
|
}, "Invoke must never panic on a truncated SSE body")
|
|
}
|
|
|
|
func TestInvoke_StreamingEmptyBody(t *testing.T) {
|
|
m := newTestMiddleware(t)
|
|
in := &middleware.Input{
|
|
Slot: middleware.SlotOnResponse,
|
|
Status: 200,
|
|
RespHeaders: []middleware.KV{{Key: "Content-Type", Value: "text/event-stream"}},
|
|
RespBody: nil,
|
|
Metadata: []middleware.KV{{Key: middleware.KeyLLMProvider, Value: "openai"}},
|
|
}
|
|
out, err := m.Invoke(context.Background(), in)
|
|
require.NoError(t, err, "empty SSE body must not surface as error")
|
|
assert.Empty(t, out.Metadata, "no metadata for empty SSE body")
|
|
}
|
|
|
|
func TestAccumulateAnthropicStream_PartialUsage(t *testing.T) {
|
|
body := []byte(`event: message_start
|
|
data: {"type":"message_start","message":{"usage":{"input_tokens":10}}}
|
|
|
|
event: content_block_delta
|
|
data: {"type":"content_block_delta","delta":{"type":"text_delta","text":"hi"}}
|
|
|
|
`)
|
|
usage, completion := accumulateAnthropicStream(body)
|
|
assert.Equal(t, int64(10), usage.InputTokens, "partial input_tokens must survive truncated stream")
|
|
assert.Equal(t, int64(0), usage.OutputTokens, "output_tokens stays zero without message_delta")
|
|
assert.Equal(t, "hi", completion, "completion must come from observed text_delta events")
|
|
}
|