FLORA

Run a one-off model generation on a project canvas. SPENDS CREDITS. Returns a run_id — poll it with flora_get_run — plus canvas_url, the link to give the user, and the node_id of the new canvas node. Pass reference_node_ids to use existing canvas nodes as image inputs; the new node is wired to them with edges. Canvas node UUIDs come from flora_list_canvas_nodes (GET /projects/{projectId}/nodes) or a previous flora_generate response's node_id. Requires an image-to-image capable model (type=image); other models are rejected with invalid_parameters. Omit model to let FLORA pick a default for the modality. Never build a project URL from an id; a prj_ id is not a route.

Get one asset's metadata, including failure_message when its upload did not finish. flora_list_assets covers the rest.

Read a project canvas as a Mermaid diagram plus node/edge counts. The diagram is a picture of the graph, not a document to edit and send back: returning it to flora_add_to_canvas duplicates every node it declares.

Get one project's metadata.

Poll a run started by flora_generate, flora_run_canvas_nodes, flora_run_technique, or flora_run_action. Returns status, progress, and output URLs once complete. Completed model3d runs also return preview images (when the provider supplied them) and a GLB resource link. This does not block — call it again after a wait. For a technique run, pass the same technique_id you ran: technique runs are addressed under their technique, and the run_id alone does not identify one.

Get one technique's declared inputs and outputs. Call this before flora_run_technique: the input ids it returns are the keys that tool's inputs map expects.

List assets in a workspace or project.

List a project canvas's media nodes with their asset URLs. Use flora_get_canvas for graph structure.

List generation history, newest first, with cost and outputs.

List the generation models available to this account. Returns a summary per model; pass model_id to get one model's full parameter list for flora_generate.

List projects in a workspace, most recently active first.

List past technique runs with their status, cost, and outputs. Also the way to find technique ids from work already done in a workspace.

List saved techniques (reusable multi-step workflows) with their declared inputs, outputs, and run cost.

List the FLORA workspaces this account can use. Start here: every other tool needs a workspace_id. If more than one comes back, name them and ask which to bill rather than taking the first — workspaces differ in plan and entitlements, so actions and project creation can succeed in one and fail in another on the same account. Pair each project_id only with its own workspace_id: a mismatch fails with 400 input_validation_error "Project does not belong to the specified workspace."

Mint a one-shot sign-in link that opens a project canvas with FLORA's in-page WebMCP tools enabled. Only for agents that have their own browser (Claude Code Desktop, Cowork, or any host with a browser/JavaScript tool): open the returned url in a tab, then drive the canvas via navigator.modelContext. The link works once and the session acts as the user for session_hours — never share it or put it in a reply. Agents without a browser should keep using the flora_* tools instead.

Run a prebuilt action headlessly on inputs you supply inline. Credit-free. This does NOT touch the canvas — nothing appears in the project, and outputs come back from flora_get_run. project_id only scopes authorization and generation history. To produce a canvas node, use flora_add_action then flora_run_canvas_action. Returns canvas_url; give the user that link rather than building one from an id.

Run an action node that already sits on a canvas, taking its inputs from the nodes wired into it and writing outputs back to canvas result nodes. Credit-free. Place the node first with flora_add_action, and connect its inputs — an action with required inputs and no incoming edges fails. Returns canvas_url; give the user that link rather than building one from an id.

Run generation nodes (image, video, text, audio) that already sit on a project canvas, using the model, prompt, parameters and wired upstream inputs each node carries — configure the node first with flora_add_to_canvas. SPENDS CREDITS immediately, with no confirmation step; quote a cost to the user first. Returns one entry per node_id: started entries carry a run_id to poll with flora_get_run, skipped entries carry a machine-readable reason (node_not_found, node_not_executable, no_model_configured, generation_already_running, …) and never fail the rest of the batch. Connected nodes run in dependency order, so a whole chain can be started in one call. For action nodes use flora_run_canvas_action instead.

Run a saved technique. SPENDS CREDITS. inputs is a map keyed by the input ids from flora_get_technique, valued with text or asset URLs; each value is sent as the type the technique declares for that input. URL-typed (media) inputs must use Flora-hosted or allowlisted media URLs. Upload external media first with flora_create_asset and pass its returned url. Returns a run_id — poll it with flora_get_run, passing this same technique_id.

Search the prebuilt action registry (deterministic media/text transforms — crop, stitch, split, colour-grade). Actions are credit-free.

Open an interactive 3D viewer for a run from flora_get_run. Read-only and credit-free: orbit, pan, zoom, reset, inspect metadata, or open the GLB. Supply technique_id for technique runs. Returns asset_id and GLB links for chaining without geometry bytes, plus up to three provider preview images when available. The model can evaluate only those image blocks, not the user's interactive view; no previews means visual quality is unknown. Pending runs can be refreshed. Never inline a GLB.