Add common catalog query layer and enricher service
Introduce an API-style data layer for the global common tracker pattern and common third party catalogs: typed filters, order fields, CursorKey, cursor-paginated Load and CountAll, plus by-id enrichment re-queue and a scoped reset/remap helper for a banner's tracker patterns. These reuse the same page.Cursor/filter/order types the GraphQL API consumes, so a future proboctl API can back them unchanged. Extract the common-pattern enrichment logic out of the worker into a CommonPatternEnricher service so it can run either from the background queue or synchronously over a known set of ids; the worker becomes a thin poller that delegates to it. Extract the LLM client and tracker-agents config wiring into pkg/agentsbuild so probod and other binaries build agents identically; probod now delegates to it. Signed-off-by: Émile Ré <emile@probo.com>
This commit is contained in:
Émile Ré
@@ -15,20 +15,17 @@
|
||||
package probod
|
||||
|
||||
import (
|
||||
"fmt"
|
||||
|
||||
"github.com/prometheus/client_golang/prometheus"
|
||||
"go.gearno.de/kit/httpclient"
|
||||
"go.gearno.de/kit/log"
|
||||
"go.opentelemetry.io/otel/trace"
|
||||
"go.probo.inc/probo/pkg/agentsbuild"
|
||||
"go.probo.inc/probo/pkg/llm"
|
||||
llmanthropic "go.probo.inc/probo/pkg/llm/anthropic"
|
||||
llmopenai "go.probo.inc/probo/pkg/llm/openai"
|
||||
)
|
||||
|
||||
// resolveAgentClient resolves the agent's effective config from defaults and
|
||||
// builds an LLM client for it. The name parameter is used in the logger and
|
||||
// in error messages.
|
||||
// in error messages. It delegates to pkg/agentsbuild so probod and proboctl
|
||||
// wire LLM clients identically.
|
||||
func (impl *Implm) resolveAgentClient(
|
||||
name string,
|
||||
agent LLMAgentConfig,
|
||||
@@ -36,61 +33,5 @@ func (impl *Implm) resolveAgentClient(
|
||||
tp trace.TracerProvider,
|
||||
r prometheus.Registerer,
|
||||
) (LLMAgentConfig, *llm.Client, error) {
|
||||
resolved := impl.cfg.Agents.ResolveAgent(agent)
|
||||
|
||||
providerCfg, ok := impl.cfg.Agents.Providers[resolved.Provider]
|
||||
if !ok {
|
||||
return LLMAgentConfig{}, nil, fmt.Errorf("unknown LLM provider %q for %s agent", resolved.Provider, name)
|
||||
}
|
||||
|
||||
client, err := buildLLMClient(providerCfg, l.Named("llm."+name), tp, r)
|
||||
if err != nil {
|
||||
return LLMAgentConfig{}, nil, fmt.Errorf("cannot create %s LLM client: %w", name, err)
|
||||
}
|
||||
|
||||
return resolved, client, nil
|
||||
}
|
||||
|
||||
func buildLLMClient(cfg LLMProviderConfig, l *log.Logger, tp trace.TracerProvider, r prometheus.Registerer) (*llm.Client, error) {
|
||||
providerType := cfg.Type
|
||||
if providerType == "" {
|
||||
providerType = "openai"
|
||||
}
|
||||
|
||||
httpClient := httpclient.DefaultPooledClient(
|
||||
httpclient.WithLogger(l),
|
||||
httpclient.WithTracerProvider(tp),
|
||||
httpclient.WithRegisterer(r),
|
||||
)
|
||||
|
||||
switch providerType {
|
||||
case "openai":
|
||||
p := llmopenai.NewProvider(
|
||||
cfg.APIKey,
|
||||
llmopenai.WithHTTPClient(httpClient),
|
||||
)
|
||||
|
||||
return llm.NewClient(
|
||||
p,
|
||||
"openai",
|
||||
llm.WithLogger(l),
|
||||
llm.WithTracerProvider(tp),
|
||||
), nil
|
||||
case "anthropic":
|
||||
p := llmanthropic.NewProvider(
|
||||
cfg.APIKey,
|
||||
llmanthropic.WithHTTPClient(httpClient),
|
||||
)
|
||||
|
||||
return llm.NewClient(
|
||||
p,
|
||||
"anthropic",
|
||||
llm.WithLogger(l),
|
||||
llm.WithTracerProvider(tp),
|
||||
), nil
|
||||
case "bedrock":
|
||||
return nil, fmt.Errorf("bedrock provider not yet wired; requires aws.Config")
|
||||
default:
|
||||
return nil, fmt.Errorf("unsupported LLM provider type: %q", providerType)
|
||||
}
|
||||
return agentsbuild.ResolveAgentClient(impl.cfg.Agents, name, agent, l, tp, r)
|
||||
}
|
||||
|
||||
@@ -15,78 +15,22 @@
|
||||
package probod
|
||||
|
||||
import (
|
||||
"fmt"
|
||||
"time"
|
||||
|
||||
"github.com/prometheus/client_golang/prometheus"
|
||||
"go.gearno.de/kit/log"
|
||||
"go.opentelemetry.io/otel/trace"
|
||||
"go.probo.inc/probo/pkg/agentsbuild"
|
||||
"go.probo.inc/probo/pkg/cookiebanner"
|
||||
"go.probo.inc/probo/pkg/thirdparty"
|
||||
)
|
||||
|
||||
// buildTrackerAgentsConfig wires the tracker agents that share one LLM
|
||||
// client and model: the tracker-mapping agent (catalog identification),
|
||||
// the common-pattern enrichment agent (description research), and the
|
||||
// third-party disambiguation agent that the tracker-mapping worker uses
|
||||
// to promote patterns to org ThirdParties. All are opt-in: deployments
|
||||
// that do not set `llm.tracker-mapping.provider` get zero configs (nil
|
||||
// LLM client) so the workers run without agent fallback.
|
||||
//
|
||||
// The agents are sourced from the same `tracker-mapping` config slot
|
||||
// because they share the LLM client, model, and lifecycle. The
|
||||
// disambiguation agent has no Firecrawl/DB tools, so its config
|
||||
// surface is narrower and it lives in the cross-domain pkg/thirdparty
|
||||
// package.
|
||||
// buildTrackerAgentsConfig wires the tracker agents (mapping, enrichment,
|
||||
// disambiguation) from the probod config. It delegates to pkg/agentsbuild
|
||||
// so probod and proboctl build the same agent configuration; see that
|
||||
// package for the wiring rationale.
|
||||
func (impl *Implm) buildTrackerAgentsConfig(
|
||||
l *log.Logger,
|
||||
tp trace.TracerProvider,
|
||||
r prometheus.Registerer,
|
||||
) (cookiebanner.TrackerAgentsConfig, thirdparty.DisambiguationConfig, error) {
|
||||
if impl.cfg.Agents.TrackerMapping.Provider == "" {
|
||||
return cookiebanner.TrackerAgentsConfig{}, thirdparty.DisambiguationConfig{}, nil
|
||||
}
|
||||
|
||||
agentCfg, llmClient, err := impl.resolveAgentClient(
|
||||
"tracker-mapping",
|
||||
impl.cfg.Agents.TrackerMapping,
|
||||
l,
|
||||
tp,
|
||||
r,
|
||||
)
|
||||
if err != nil {
|
||||
return cookiebanner.TrackerAgentsConfig{}, thirdparty.DisambiguationConfig{}, fmt.Errorf("cannot resolve tracker mapping agent client: %w", err)
|
||||
}
|
||||
|
||||
mappingWorkerCfg := impl.cfg.TrackerMappingWorker
|
||||
enrichmentWorkerCfg := impl.cfg.CommonPatternEnrichmentWorker
|
||||
|
||||
// The mapping and enrichment agents share one config slot but run
|
||||
// from separate workers with separate max-turns. AgentTimeout here
|
||||
// carries the mapping worker's value (also reused by the
|
||||
// disambiguation agent); the enrichment worker overrides it on its
|
||||
// own copy at registration.
|
||||
trackerAgentsCfg := cookiebanner.TrackerAgentsConfig{
|
||||
LLMClient: llmClient,
|
||||
Model: agentCfg.ModelName,
|
||||
FirecrawlAPIKey: impl.cfg.Agents.Tools.FirecrawlAPIKey,
|
||||
MaxTokens: agentCfg.MaxTokens,
|
||||
Temperature: agentCfg.Temperature,
|
||||
AgentTimeout: time.Duration(mappingWorkerCfg.AgentTimeout) * time.Second,
|
||||
MappingMaxTurns: mappingWorkerCfg.AgentMaxTurns,
|
||||
EnrichmentMaxTurns: enrichmentWorkerCfg.AgentMaxTurns,
|
||||
}
|
||||
|
||||
// The disambiguation agent emits a single id plus a short rationale,
|
||||
// so it keeps its own smaller token budget (left unset here) rather
|
||||
// than inheriting the mapping agent's. It shares the mapping worker's
|
||||
// timeout.
|
||||
disambiguationCfg := thirdparty.DisambiguationConfig{
|
||||
LLMClient: llmClient,
|
||||
Model: agentCfg.ModelName,
|
||||
Temperature: agentCfg.Temperature,
|
||||
Timeout: time.Duration(mappingWorkerCfg.AgentTimeout) * time.Second,
|
||||
}
|
||||
|
||||
return trackerAgentsCfg, disambiguationCfg, nil
|
||||
return agentsbuild.BuildTrackerAgentsConfig(impl.cfg, l, tp, r)
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user