refactor: replace fmt.Errorf and errors.New with coreerr.E from go-log

Replace all remaining fmt.Errorf and errors.New calls in production code
with structured coreerr.E(op, msg, err) from forge.lthn.ai/core/go-log.
Covers 10 files across the infra package and cmd/prod and cmd/monitor.

Co-Authored-By: Virgil <virgil@lethean.io>

client.go

@@ -10,6 +10,8 @@ import (
 	"strconv"
 	"sync"
 	"time"
+
+	coreerr "forge.lthn.ai/core/go-log"
 )
 
 // RetryConfig controls exponential backoff retry behaviour.
@@ -105,7 +107,7 @@ func (a *APIClient) Do(req *http.Request, result any) error {
 
 		resp, err := a.client.Do(req)
 		if err != nil {
-			lastErr = fmt.Errorf("%s: %w", a.prefix, err)
+			lastErr = coreerr.E(a.prefix, "request failed", err)
 			if attempt < attempts-1 {
 				a.backoff(attempt, req)
 			}
@@ -115,7 +117,7 @@ func (a *APIClient) Do(req *http.Request, result any) error {
 		data, err := io.ReadAll(resp.Body)
 		_ = resp.Body.Close()
 		if err != nil {
-			lastErr = fmt.Errorf("read response: %w", err)
+			lastErr = coreerr.E("client.Do", "read response", err)
 			if attempt < attempts-1 {
 				a.backoff(attempt, req)
 			}
@@ -129,7 +131,7 @@ func (a *APIClient) Do(req *http.Request, result any) error {
 			a.blockedUntil = time.Now().Add(retryAfter)
 			a.mu.Unlock()
 
-			lastErr = fmt.Errorf("%s %d: rate limited", a.prefix, resp.StatusCode)
+			lastErr = coreerr.E(a.prefix, fmt.Sprintf("rate limited: HTTP %d", resp.StatusCode), nil)
 			if attempt < attempts-1 {
 				select {
 				case <-req.Context().Done():
@@ -142,7 +144,7 @@ func (a *APIClient) Do(req *http.Request, result any) error {
 
 		// Server errors are retryable.
 		if resp.StatusCode >= 500 {
-			lastErr = fmt.Errorf("%s %d: %s", a.prefix, resp.StatusCode, string(data))
+			lastErr = coreerr.E(a.prefix, fmt.Sprintf("HTTP %d: %s", resp.StatusCode, string(data)), nil)
 			if attempt < attempts-1 {
 				a.backoff(attempt, req)
 			}
@@ -151,13 +153,13 @@ func (a *APIClient) Do(req *http.Request, result any) error {
 
 		// Client errors (4xx, except 429 handled above) are not retried.
 		if resp.StatusCode >= 400 {
-			return fmt.Errorf("%s %d: %s", a.prefix, resp.StatusCode, string(data))
+			return coreerr.E(a.prefix, fmt.Sprintf("HTTP %d: %s", resp.StatusCode, string(data)), nil)
 		}
 
 		// Success — decode if requested.
 		if result != nil {
 			if err := json.Unmarshal(data, result); err != nil {
-				return fmt.Errorf("decode response: %w", err)
+				return coreerr.E("client.Do", "decode response", err)
 			}
 		}
 		return nil
@@ -191,7 +193,7 @@ func (a *APIClient) DoRaw(req *http.Request) ([]byte, error) {
 
 		resp, err := a.client.Do(req)
 		if err != nil {
-			lastErr = fmt.Errorf("%s: %w", a.prefix, err)
+			lastErr = coreerr.E(a.prefix, "request failed", err)
 			if attempt < attempts-1 {
 				a.backoff(attempt, req)
 			}
@@ -201,7 +203,7 @@ func (a *APIClient) DoRaw(req *http.Request) ([]byte, error) {
 		data, err := io.ReadAll(resp.Body)
 		_ = resp.Body.Close()
 		if err != nil {
-			lastErr = fmt.Errorf("read response: %w", err)
+			lastErr = coreerr.E("client.DoRaw", "read response", err)
 			if attempt < attempts-1 {
 				a.backoff(attempt, req)
 			}
@@ -214,7 +216,7 @@ func (a *APIClient) DoRaw(req *http.Request) ([]byte, error) {
 			a.blockedUntil = time.Now().Add(retryAfter)
 			a.mu.Unlock()
 
-			lastErr = fmt.Errorf("%s %d: rate limited", a.prefix, resp.StatusCode)
+			lastErr = coreerr.E(a.prefix, fmt.Sprintf("rate limited: HTTP %d", resp.StatusCode), nil)
 			if attempt < attempts-1 {
 				select {
 				case <-req.Context().Done():
@@ -226,7 +228,7 @@ func (a *APIClient) DoRaw(req *http.Request) ([]byte, error) {
 		}
 
 		if resp.StatusCode >= 500 {
-			lastErr = fmt.Errorf("%s %d: %s", a.prefix, resp.StatusCode, string(data))
+			lastErr = coreerr.E(a.prefix, fmt.Sprintf("HTTP %d: %s", resp.StatusCode, string(data)), nil)
 			if attempt < attempts-1 {
 				a.backoff(attempt, req)
 			}
@@ -234,7 +236,7 @@ func (a *APIClient) DoRaw(req *http.Request) ([]byte, error) {
 		}
 
 		if resp.StatusCode >= 400 {
-			return nil, fmt.Errorf("%s %d: %s", a.prefix, resp.StatusCode, string(data))
+			return nil, coreerr.E(a.prefix, fmt.Sprintf("HTTP %d: %s", resp.StatusCode, string(data)), nil)
 		}
 
 		return data, nil

cloudns.go

@@ -3,10 +3,11 @@ package infra
 import (
 	"context"
 	"encoding/json"
-	"fmt"
 	"net/http"
 	"net/url"
 	"strconv"
+
+	coreerr "forge.lthn.ai/core/go-log"
 )
 
 const cloudnsBaseURL = "https://api.cloudns.net"
@@ -81,7 +82,7 @@ func (c *CloudNSClient) ListRecords(ctx context.Context, domain string) (map[str
 
 	var records map[string]CloudNSRecord
 	if err := json.Unmarshal(data, &records); err != nil {
-		return nil, fmt.Errorf("parse records: %w", err)
+		return nil, coreerr.E("CloudNSClient.ListRecords", "parse records", err)
 	}
 	return records, nil
 }
@@ -108,11 +109,11 @@ func (c *CloudNSClient) CreateRecord(ctx context.Context, domain, host, recordTy
 		} `json:"data"`
 	}
 	if err := json.Unmarshal(data, &result); err != nil {
-		return "", fmt.Errorf("parse response: %w", err)
+		return "", coreerr.E("CloudNSClient.CreateRecord", "parse response", err)
 	}
 
 	if result.Status != "Success" {
-		return "", fmt.Errorf("cloudns: %s", result.StatusDescription)
+		return "", coreerr.E("CloudNSClient.CreateRecord", result.StatusDescription, nil)
 	}
 
 	return strconv.Itoa(result.Data.ID), nil
@@ -138,11 +139,11 @@ func (c *CloudNSClient) UpdateRecord(ctx context.Context, domain, recordID, host
 		StatusDescription string `json:"statusDescription"`
 	}
 	if err := json.Unmarshal(data, &result); err != nil {
-		return fmt.Errorf("parse response: %w", err)
+		return coreerr.E("CloudNSClient.UpdateRecord", "parse response", err)
 	}
 
 	if result.Status != "Success" {
-		return fmt.Errorf("cloudns: %s", result.StatusDescription)
+		return coreerr.E("CloudNSClient.UpdateRecord", result.StatusDescription, nil)
 	}
 
 	return nil
@@ -164,11 +165,11 @@ func (c *CloudNSClient) DeleteRecord(ctx context.Context, domain, recordID strin
 		StatusDescription string `json:"statusDescription"`
 	}
 	if err := json.Unmarshal(data, &result); err != nil {
-		return fmt.Errorf("parse response: %w", err)
+		return coreerr.E("CloudNSClient.DeleteRecord", "parse response", err)
 	}
 
 	if result.Status != "Success" {
-		return fmt.Errorf("cloudns: %s", result.StatusDescription)
+		return coreerr.E("CloudNSClient.DeleteRecord", result.StatusDescription, nil)
 	}
 
 	return nil
@@ -179,7 +180,7 @@ func (c *CloudNSClient) DeleteRecord(ctx context.Context, domain, recordID strin
 func (c *CloudNSClient) EnsureRecord(ctx context.Context, domain, host, recordType, value string, ttl int) (bool, error) {
 	records, err := c.ListRecords(ctx, domain)
 	if err != nil {
-		return false, fmt.Errorf("list records: %w", err)
+		return false, coreerr.E("CloudNSClient.EnsureRecord", "list records", err)
 	}
 
 	// Check if record already exists
@@ -190,7 +191,7 @@ func (c *CloudNSClient) EnsureRecord(ctx context.Context, domain, host, recordTy
 			}
 			// Update existing record
 			if err := c.UpdateRecord(ctx, domain, id, host, recordType, value, ttl); err != nil {
-				return false, fmt.Errorf("update record: %w", err)
+				return false, coreerr.E("CloudNSClient.EnsureRecord", "update record", err)
 			}
 			return true, nil
 		}
@@ -198,7 +199,7 @@ func (c *CloudNSClient) EnsureRecord(ctx context.Context, domain, host, recordTy
 
 	// Create new record
 	if _, err := c.CreateRecord(ctx, domain, host, recordType, value, ttl); err != nil {
-		return false, fmt.Errorf("create record: %w", err)
+		return false, coreerr.E("CloudNSClient.EnsureRecord", "create record", err)
 	}
 	return true, nil
 }

cmd/monitor/cmd_monitor.go

@@ -582,5 +582,5 @@ func parseGitHubRepo(url string) (string, error) {
 		}
 	}
 
-	return "", fmt.Errorf("could not parse GitHub repo from URL: %s", url)
+	return "", log.E("monitor.parseGitHubRepo", "could not parse GitHub repo from URL: "+url, nil)
 }

cmd/prod/cmd_dns.go

@@ -2,12 +2,11 @@ package prod
 
 import (
 	"context"
-	"errors"
-	"fmt"
 	"os"
 	"time"
 
 	"forge.lthn.ai/core/cli/pkg/cli"
+	coreerr "forge.lthn.ai/core/go-log"
 	"forge.lthn.ai/core/go-infra"
 )
 
@@ -56,7 +55,7 @@ func getDNSClient() (*infra.CloudNSClient, error) {
 	authID := os.Getenv("CLOUDNS_AUTH_ID")
 	authPass := os.Getenv("CLOUDNS_AUTH_PASSWORD")
 	if authID == "" || authPass == "" {
-		return nil, errors.New("CLOUDNS_AUTH_ID and CLOUDNS_AUTH_PASSWORD required")
+		return nil, coreerr.E("prod.getDNSClient", "CLOUDNS_AUTH_ID and CLOUDNS_AUTH_PASSWORD required", nil)
 	}
 	return infra.NewCloudNSClient(authID, authPass), nil
 }
@@ -77,7 +76,7 @@ func runDNSList(cmd *cli.Command, args []string) error {
 
 	records, err := dns.ListRecords(ctx, zone)
 	if err != nil {
-		return fmt.Errorf("list records: %w", err)
+		return coreerr.E("prod.runDNSList", "list records", err)
 	}
 
 	cli.Print("%s DNS records for %s\n\n", cli.BoldStyle.Render("▶"), cli.TitleStyle.Render(zone))
@@ -114,7 +113,7 @@ func runDNSSet(cmd *cli.Command, args []string) error {
 
 	changed, err := dns.EnsureRecord(ctx, dnsZone, host, recordType, value, dnsTTL)
 	if err != nil {
-		return fmt.Errorf("set record: %w", err)
+		return coreerr.E("prod.runDNSSet", "set record", err)
 	}
 
 	if changed {

cmd/prod/cmd_lb.go

@@ -2,12 +2,12 @@ package prod
 
 import (
 	"context"
-	"errors"
 	"fmt"
 	"os"
 	"time"
 
 	"forge.lthn.ai/core/cli/pkg/cli"
+	coreerr "forge.lthn.ai/core/go-log"
 	"forge.lthn.ai/core/go-infra"
 )
 
@@ -39,7 +39,7 @@ func init() {
 func getHCloudClient() (*infra.HCloudClient, error) {
 	token := os.Getenv("HCLOUD_TOKEN")
 	if token == "" {
-		return nil, errors.New("HCLOUD_TOKEN environment variable required")
+		return nil, coreerr.E("prod.getHCloudClient", "HCLOUD_TOKEN environment variable required", nil)
 	}
 	return infra.NewHCloudClient(token), nil
 }
@@ -55,7 +55,7 @@ func runLBStatus(cmd *cli.Command, args []string) error {
 
 	lbs, err := hc.ListLoadBalancers(ctx)
 	if err != nil {
-		return fmt.Errorf("list load balancers: %w", err)
+		return coreerr.E("prod.runLBStatus", "list load balancers", err)
 	}
 
 	if len(lbs) == 0 {

cmd/prod/cmd_setup.go

@@ -2,12 +2,11 @@ package prod
 
 import (
 	"context"
-	"errors"
-	"fmt"
 	"os"
 	"time"
 
 	"forge.lthn.ai/core/cli/pkg/cli"
+	coreerr "forge.lthn.ai/core/go-log"
 	"forge.lthn.ai/core/go-infra"
 )
 
@@ -71,7 +70,7 @@ func runSetup(cmd *cli.Command, args []string) error {
 
 		if err := step.fn(ctx, cfg); err != nil {
 			cli.Print("  %s %s: %s\n", cli.ErrorStyle.Render("✗"), step.name, err)
-			return fmt.Errorf("step %s failed: %w", step.name, err)
+			return coreerr.E("prod.setup", "step "+step.name+" failed", err)
 		}
 
 		cli.Print("  %s %s complete\n", cli.SuccessStyle.Render("✓"), step.name)
@@ -90,7 +89,7 @@ func stepDiscover(ctx context.Context, cfg *infra.Config) error {
 		hc := infra.NewHCloudClient(hcloudToken)
 		servers, err := hc.ListServers(ctx)
 		if err != nil {
-			return fmt.Errorf("list HCloud servers: %w", err)
+			return coreerr.E("prod.stepDiscover", "list HCloud servers", err)
 		}
 
 		for _, s := range servers {
@@ -115,7 +114,7 @@ func stepDiscover(ctx context.Context, cfg *infra.Config) error {
 		hr := infra.NewHRobotClient(robotUser, robotPass)
 		servers, err := hr.ListServers(ctx)
 		if err != nil {
-			return fmt.Errorf("list Robot servers: %w", err)
+			return coreerr.E("prod.stepDiscover", "list Robot servers", err)
 		}
 
 		for _, s := range servers {
@@ -141,7 +140,7 @@ func stepDiscover(ctx context.Context, cfg *infra.Config) error {
 func stepLoadBalancer(ctx context.Context, cfg *infra.Config) error {
 	hcloudToken := os.Getenv("HCLOUD_TOKEN")
 	if hcloudToken == "" {
-		return errors.New("HCLOUD_TOKEN required for load balancer management")
+		return coreerr.E("prod.stepLoadBalancer", "HCLOUD_TOKEN required for load balancer management", nil)
 	}
 
 	hc := infra.NewHCloudClient(hcloudToken)
@@ -149,7 +148,7 @@ func stepLoadBalancer(ctx context.Context, cfg *infra.Config) error {
 	// Check if LB already exists
 	lbs, err := hc.ListLoadBalancers(ctx)
 	if err != nil {
-		return fmt.Errorf("list load balancers: %w", err)
+		return coreerr.E("prod.stepLoadBalancer", "list load balancers", err)
 	}
 
 	for _, lb := range lbs {
@@ -176,7 +175,7 @@ func stepLoadBalancer(ctx context.Context, cfg *infra.Config) error {
 	for _, b := range cfg.LoadBalancer.Backends {
 		host, ok := cfg.Hosts[b.Host]
 		if !ok {
-			return fmt.Errorf("backend host '%s' not found in config", b.Host)
+			return coreerr.E("prod.stepLoadBalancer", "backend host '"+b.Host+"' not found in config", nil)
 		}
 		targets = append(targets, infra.HCloudLBCreateTarget{
 			Type: "ip",
@@ -224,7 +223,7 @@ func stepLoadBalancer(ctx context.Context, cfg *infra.Config) error {
 
 	lb, err := hc.CreateLoadBalancer(ctx, req)
 	if err != nil {
-		return fmt.Errorf("create load balancer: %w", err)
+		return coreerr.E("prod.stepLoadBalancer", "create load balancer", err)
 	}
 
 	cli.Print("  Created: %s (ID: %d, IP: %s)\n",
@@ -237,7 +236,7 @@ func stepDNS(ctx context.Context, cfg *infra.Config) error {
 	authID := os.Getenv("CLOUDNS_AUTH_ID")
 	authPass := os.Getenv("CLOUDNS_AUTH_PASSWORD")
 	if authID == "" || authPass == "" {
-		return errors.New("CLOUDNS_AUTH_ID and CLOUDNS_AUTH_PASSWORD required")
+		return coreerr.E("prod.stepDNS", "CLOUDNS_AUTH_ID and CLOUDNS_AUTH_PASSWORD required", nil)
 	}
 
 	dns := infra.NewCloudNSClient(authID, authPass)

cmd/prod/cmd_ssh.go

@@ -7,6 +7,7 @@ import (
 	"syscall"
 
 	"forge.lthn.ai/core/cli/pkg/cli"
+	coreerr "forge.lthn.ai/core/go-log"
 )
 
 var sshCmd = &cli.Command{
@@ -37,7 +38,7 @@ func runSSH(cmd *cli.Command, args []string) error {
 		for n, h := range cfg.Hosts {
 			cli.Print("  %s  %s  (%s)\n", cli.BoldStyle.Render(n), h.IP, h.Role)
 		}
-		return fmt.Errorf("host '%s' not found in infra.yaml", name)
+		return coreerr.E("prod.ssh", "host '"+name+"' not found in infra.yaml", nil)
 	}
 
 	sshArgs := []string{
@@ -55,7 +56,7 @@ func runSSH(cmd *cli.Command, args []string) error {
 
 	sshPath, err := exec.LookPath("ssh")
 	if err != nil {
-		return fmt.Errorf("ssh not found: %w", err)
+		return coreerr.E("prod.ssh", "ssh not found", err)
 	}
 
 	// Replace current process with SSH

cmd/prod/cmd_status.go

@@ -10,6 +10,7 @@ import (
 
 	"forge.lthn.ai/core/go-ansible"
 	"forge.lthn.ai/core/cli/pkg/cli"
+	coreerr "forge.lthn.ai/core/go-log"
 	"forge.lthn.ai/core/go-infra"
 )
 
@@ -114,14 +115,14 @@ func checkHost(ctx context.Context, name string, host *infra.Host) hostStatus {
 
 	client, err := ansible.NewSSHClient(sshCfg)
 	if err != nil {
-		s.Error = fmt.Errorf("create SSH client: %w", err)
+		s.Error = coreerr.E("prod.checkHost", "create SSH client", err)
 		return s
 	}
 	defer func() { _ = client.Close() }()
 
 	start := time.Now()
 	if err := client.Connect(ctx); err != nil {
-		s.Error = fmt.Errorf("SSH connect: %w", err)
+		s.Error = coreerr.E("prod.checkHost", "SSH connect", err)
 		return s
 	}
 	s.Connected = true

config.go

@@ -3,10 +3,10 @@
 package infra
 
 import (
-	"fmt"
 	"os"
 	"path/filepath"
 
+	coreerr "forge.lthn.ai/core/go-log"
 	coreio "forge.lthn.ai/core/go-io"
 	"gopkg.in/yaml.v3"
 )
@@ -232,12 +232,12 @@ type BackupJob struct {
 func Load(path string) (*Config, error) {
 	data, err := coreio.Local.Read(path)
 	if err != nil {
-		return nil, fmt.Errorf("read infra config: %w", err)
+		return nil, coreerr.E("infra.Load", "read infra config", err)
 	}
 
 	var cfg Config
 	if err := yaml.Unmarshal([]byte(data), &cfg); err != nil {
-		return nil, fmt.Errorf("parse infra config: %w", err)
+		return nil, coreerr.E("infra.Load", "parse infra config", err)
 	}
 
 	// Expand SSH key paths
@@ -269,7 +269,7 @@ func Discover(startDir string) (*Config, string, error) {
 		}
 		dir = parent
 	}
-	return nil, "", fmt.Errorf("infra.yaml not found (searched from %s)", startDir)
+	return nil, "", coreerr.E("infra.Discover", "infra.yaml not found (searched from "+startDir+")", nil)
 }
 
 // HostsByRole returns all hosts matching the given role.

hetzner.go

@@ -6,6 +6,8 @@ import (
 	"fmt"
 	"net/http"
 	"strings"
+
+	coreerr "forge.lthn.ai/core/go-log"
 )
 
 const (
@@ -210,7 +212,7 @@ func (c *HCloudClient) GetLoadBalancer(ctx context.Context, id int) (*HCloudLoad
 func (c *HCloudClient) CreateLoadBalancer(ctx context.Context, req HCloudLBCreateRequest) (*HCloudLoadBalancer, error) {
 	body, err := json.Marshal(req)
 	if err != nil {
-		return nil, fmt.Errorf("marshal request: %w", err)
+		return nil, coreerr.E("HCloudClient.CreateLoadBalancer", "marshal request", err)
 	}
 
 	var result struct {