What causes "%s: refresh card: %w" in chenhg5/cc-connect?

Check network connectivity to open.feishu.cn (or the configured domain) and retry the operation.

%s: refresh card: %w

Error message

%s: refresh card: %w

What it means

Wraps a transport-level error from the Lark SDK's Im.Message.Patch call, which is used by the Feishu platform to refresh (update) an already-sent interactive card. The original SDK error (network failure, auth, invalid message_id) is preserved via %w. It is retried transiently via withTransientRetry/withFreshTenantAccessTokenRetry before surfacing.

Source

Thrown at platform/feishu/card.go:78

	msgID := p.cardActionMsgIDs[sessionKey]
	p.cardActionMsgMu.Unlock()

	if msgID == "" {
		return fmt.Errorf("%s: no tracked card messageID for session %q", p.tag(), sessionKey)
	}

	cardJSON := renderCard(card, sessionKey)
	req := larkim.NewPatchMessageReqBuilder().
		MessageId(msgID).
		Body(larkim.NewPatchMessageReqBodyBuilder().
			Content(cardJSON).
			Build()).
		Build()
	return p.withTransientRetry(ctx, "refresh card", func() error {
		return p.withFreshTenantAccessTokenRetry(ctx, "refresh card", func(client *lark.Client, options ...larkcore.RequestOptionFunc) error {
			resp, err := client.Im.Message.Patch(ctx, req, options...)
			if err != nil {
				return fmt.Errorf("%s: refresh card: %w", p.tag(), err)
			}
			if !resp.Success() {
				return fmt.Errorf("%s: refresh card code=%d msg=%s", p.tag(), resp.Code, resp.Msg)
			}
			return nil
		})
	})
}

// renderCardMap converts a core.Card into the Feishu Interactive Card map
// using the v1 format. Used both for message API (via renderCard) and
// callback responses (CardActionTriggerResponse).
func renderCardMap(card *core.Card, sessionKey string) map[string]any {
	result := map[string]any{
		"config": map[string]any{
			"wide_screen_mode": true,
		},
	}

View on GitHub (pinned to 4000b2338a)

Solutions

Check network connectivity to open.feishu.cn (or the configured domain) and retry the operation.
Verify app_id/app_secret are valid and the token refresh path works (run cc-connect doctor).
Confirm the message_id being patched still exists (not recalled or deleted).
Inspect the wrapped SDK error (%w chain) for the underlying Feishu error code.

Defensive patterns

Strategy: retry

Try / catch

// caller
if err := p.refreshCard(ctx, req); err != nil {
    slog.Warn("feishu: card refresh failed; card may be stale", "err", err)
    // non-fatal: keep session running, stale card degrades gracefully
}

Prevention

Treat card refresh failures as non-fatal and degrade to text updates.
Monitor Feishu API status and token refresh logs.
Don't patch recalled messages; track recall state first.

When it happens

Trigger: Calling the card refresh path (progress/streaming card updates) when client.Im.Message.Patch returns err: network outage, expired/invalid tenant_access_token after retries, invalid message_id, or card schema rejected at transport level.

Common situations: Feishu API outage or rate limiting during long streaming sessions; message recalled/deleted so Patch target no longer exists; app credentials rotated mid-run causing token fetch failures.

Understand the failure class

Background: "API request failed": what wrapped HTTP errors from external APIs mean and how to find the real cause — this error's family across 29 libraries.

Related errors

AI-assisted analysis of chenhg5/cc-connect@4000b2338a (2026-09-06). Data as JSON: /api/errors/b31d937a58674c71. Report an issue: GitHub.