WiseWand

Generate ranking articles in 83 languages with WiseWand, the French AI SEO article generator, and publish them onward. Agents route every draft to an editor before it goes live under your byline.

18 actions API key available

WiseWand website ↗ Platform Documentation ↗ Capability data verified 2026-08-01

An approved subject and target keyword enter the localization content queue

Agent calls Estimate Article Cost, which spends nothing, and checks it against the batch budget

Agent calls Create Article with the project brief and persona applied, then polls Get Article Output

Agent reads the finished text and the generation history once the output is populated

Agent resolves the output language against the reviewers who actually read that language

Named locale reviewer receives the draft, its language and the connection it would publish to

Locale reviewer decides whether the publish state moves from draft to live

What This Integration Enables

WiseWand is the only content generator in this category that can finish the job by itself. Create Article turns a subject into a full SEO article in any of 83 output languages, and Publish To WordPress will put that article on a live site in the same flow, with no person having read a word of it. Every publishing action accepts draft, published or scheduled, and which one a flow chooses is the entire safety design. This connector treats that parameter as the decision it is rather than a configuration detail.

The rest of the surface is built for programs rather than one-off pieces. Estimate Article Cost prices a run and spends nothing, which is the right guard in front of a bulk generation batch. Projects carry a content brief that supplies default tone, audience and settings, and personas reproduce a consistent voice across articles, so a program keeps sounding like one publisher instead of eighty. Generation is asynchronous throughout: Create Article returns an ID immediately and the flow polls Get Article Output until the text arrives, with Run Article regenerating against saved settings while the previous versions stay in the output history. Publishing reaches WordPress, Shopify, WooCommerce, PrestaShop and a webhook target, and List Publish Errors exists because articles that generate cleanly can still fail to land, which is exactly the kind of silent gap that should surface as one of the automation exceptions a flow tracks.

Without FlowRunner

Cost discovered mid-batch A bulk generation run stops halfway when credits run out, with no estimate taken beforehand

Voice reset on every article Tone and audience are re-specified per piece, so output drifts between writers and weeks

Failed publishes go unnoticed An article generates successfully, fails to reach the store, and nobody finds out until the slot is empty

With FlowRunner

Budget checked before generation Estimate Article Cost prices the run against the remaining balance before a single credit is spent

Voice carried by project and persona A saved persona and a project brief apply the same tone and audience to every article in the program

Delivery failures surfaced as exceptions List Publish Errors is read after every run and unresolved failures reach an owner with the article attached

Use Case Scenarios

A Localized Program With a Reviewer Per Language

A store sells into six markets and the content plan calls for the same twelve subjects in each. The agent runs Estimate Article Cost across the full matrix first and posts the projected credit spend for approval. It then calls Create Article per subject and language with the market's project brief and persona applied, polls Get Article Output until each is populated, and routes every finished draft to the reviewer who reads that language. Nothing publishes on the agent's own authority. The English drafts go to the content lead, the German drafts go to the German reviewer, and any language with no reviewer assigned stays in draft and gets raised rather than shipped.

Product Content Into a Store, Held at Draft

A new category launches in Shopify. The agent creates the supporting articles with the store's persona applied for voice, then calls Publish To Shopify with the status set to draft and the target blog resolved from List Connections. The merchandiser opens the store's own editor, reads the drafts in place, adjusts the product references, and flips them live. The agent then reads List Publish Errors to confirm nothing failed silently between WiseWand and the store, and posts a summary to Slack with any unresolved failures named.

Regenerating a Piece That Came Back Weak

An editor rejects a draft as too generic. Rather than starting over, the agent updates the article's configuration with a tighter brief and a different persona, calls Run Article to regenerate against the saved settings, and polls Get Article Output for the new version. Because the output history keeps every previous generation, the editor can compare the two side by side and pick, which is a better conversation than a binary approve or reject. Only the version a person chose is passed to a publishing action.

Human-in-Loop Highlight

The hard case in a multilingual program is not whether a human reviews the draft, it is whether the human reviewing it can read it. WiseWand generates in 83 languages, and a content lead who confidently approves the English article has no way to judge whether the Polish one is fluent, on-brand, or accidentally claiming something the legal team would never sign off on. A flow that sends every locale to the same approver is a rubber stamp with extra steps. So the agent resolves the language before it resolves the approver, and the message names both: "Article ready: [subject], language [language], generated with persona [name] under project [name]. Currently a draft. Publishing target is [connection], which is a live [platform] site. Routing to [reviewer] as the assigned reader for this language. Approve to publish, request a regeneration, or hold?" When no reviewer is assigned for that language, the agent does not fall back to the nearest available person. It leaves the article in draft and raises the gap, because an unread article on a live domain is the outcome the whole design exists to prevent.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via preferred channel

Human decides

Agent resumes with decision

Agent Capabilities

18 actions

Articles

Create Article Creates an SEO article and starts generating it, returning the article ID immediately. Used as the entry point for any content program, with optional keywords, length, language, country, project brief, persona, FAQ, table of contents, cover image, internal links, embedded video, call to action and schema.org markup.
Estimate Article Cost Calculates the credits an article with the given settings would consume and reports the remaining balance, without generating anything or spending credits. Used to gate a bulk run behind a budget check instead of discovering the ceiling halfway through a batch.
List Articles Returns a paginated list of workspace articles, filterable by search text, generation status, project, persona, author, category and publication state. Used to find article IDs and to build an editorial view of what is generated, published or still waiting.
Get Article Retrieves a single article record with its configuration and current generation status. Used to check settings or progress when the finished text is not what the flow needs yet.
Get Article Output Retrieves the generated content of an article along with the history of every previous generation. Used as the polling step after Create Article or Run Article, and as the source of the side-by-side comparison when an editor is choosing between versions.
Run Article Regenerates an existing article from its saved configuration, producing a new output while keeping earlier versions in the history. Used after a rejection or a settings change, so a weak draft becomes an iteration rather than a restart.

Publishing

Publish To WordPress Publishes a generated article to a connected WordPress site as a draft, a live post or a scheduled post, with optional author, categories and scheduled date. Used at the end of an approved editorial flow, and the action whose status parameter decides whether a person reads the piece first.
Publish To Shopify Publishes a generated article to a connected Shopify blog as a draft, a live post or a scheduled post. Used for store content programs where the merchandiser reviews in the store's own editor before going live.
Publish To WooCommerce Sends a generated article to a connected WooCommerce store. Used to move approved content into a WordPress commerce site through the connection configured in WiseWand.
Publish To PrestaShop Sends a generated article to a connected PrestaShop store. Used for the same approved-content handoff on PrestaShop storefronts.
Publish To Webhook Posts a generated article to a webhook connection configured in WiseWand so a custom system can receive it. Used for headless setups and internal review tools, delivering one article per call rather than creating a subscription.
List Publish Errors Lists publication failures that have not yet been resolved. Used after every publishing run, since an article can generate perfectly and still never reach the target site, which is the failure mode nobody notices until a slot is empty.

Projects, Voice and Connections

List Projects Returns a paginated list of projects, each carrying a content brief that supplies default tone, audience and settings to the articles created inside it. Used to keep a program consistent without re-specifying the brief per article.
List Personas Returns the saved writing personas in the workspace. Used to apply a consistent brand voice across a content program and to record which voice produced a given draft.
List Connections Returns the publishing connections configured in the workspace across WordPress, Shopify, WooCommerce, PrestaShop and webhook targets. Used to resolve the target before a publish action and to state in an approval message exactly which live site is involved.

Account

List Transactions Lists credit transactions over a date range, optionally restricted to debits or a single reason. Used to reconcile content spend against output produced, per project or per month.
Get Entity Summary Returns counts of the content entities in the workspace, such as articles, category pages and product pages. Used as a lightweight program dashboard metric and as a health check that the key is working.
Get Job Status Retrieves the status of a background job by ID. Used to monitor long-running operations such as bulk generation batches without polling every article individually.

Frequently Asked Questions

What can FlowRunner do with WiseWand?

FlowRunner agents can run Create Article, Estimate Article Cost, and List Articles in WiseWand, plus 15 more actions.

Does connecting WiseWand to FlowRunner require OAuth?

No. WiseWand connects to FlowRunner with an API key, no OAuth flow required.

Can WiseWand trigger a FlowRunner workflow automatically?

WiseWand doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.

Start building with WiseWand

$100 in credits. No card required. Connect in minutes.

Start Building Free Talk to us