genable-mcp 0.1.0

package/dist/tools-schema.json

@@ -0,0 +1,1220 @@
+[
+  {
+    "name": "jsx",
+    "description": "Create design trees with nested JSX markup. One jsx call builds a complete subtree atomically — nesting is the hierarchy. Keep a single logical unit inside one call; the returned root's children are already built, not stubs to be filled in later.\n\nExamples:\n  jsx({markup: \"<frame name='Card' layout='column' padding={16} fill='#FFFFFF' w='fill' />\"})\n  jsx({markup: \"<frame name='Row' layout='row' gap={8} padding={12} w='fill'><icon name='lucide:settings' size={20} /><text name='Label' w='fill'>Account</text><icon name='lucide:chevron-right' size={16} /></frame>\"})\n\nElements: frame, text, rect, ellipse, line, icon, image, instance, component, group, section, vector\nAttributes (frame): layout, justify, items, wrap, w, h, minW, maxW, p, gap, bg, fill, rounded, stroke, shadow, blur, bgblur, opacity, layoutPositioning\nAttributes (text): size, weight, lineHeight, font, fill, w (w=\"fill\" for wrap), maxLines, textTruncation\nEffects: shadow=\"0,8,32,0,#0006\" or shadow={shadow(0,8,32,0,'#0006')}; blur={10} for layer blur; bgblur={20} for frosted-glass/glassmorphism background blur. Multiple effects merge automatically.\nDecoration in auto-layout: floating orbs/blobs/decorative shapes inside a row/column parent need layoutPositioning=\"absolute\" so they don't get stacked into the main-axis flow.\nFull-frame backgrounds: set the parent frame's bg directly (supports gradients via bg=\"linear-gradient(135deg, #A 0%, #B 100%)\"). Don't add a separate <rect> backdrop.\nText: <text size={24}>content here</text>\nInstance: <instance ref=\"Button\" variant=\"Size=Large\"/>\nSelf-closing: <line w=\"fill\" stroke=\"#E5E7EB\"/> (use line for dividers/separators; rect/ellipse for SMALL pure decoration with no children — page-level backgrounds belong on the parent frame's bg)\nArc/Ring: <ellipse w={120} h={120} arc=\"0 270\" fill=\"#4F46E5\"/> (arc=\"start end innerRadius?\" — innerRadius 0-1 makes a donut/ring)\nGrid layout: call help({ name: \"grid-layout\" }) for tracks, gaps, and when row/column is a better fit\nVariable binding: fill/bg/stroke accept qualified bare-name token strings (e.g. bg=\"$Theme/Bg/Surface\"). Object literals (fill={{variable_id:...}}) drop the binding silently — always use the string form.\n\nSwap an existing subtree: jsx({replaceId: \"<id>\", markup: \"...\"}) replaces the old node at the same parent and sibling index atomically, preserving position in one call. Markup must have a single root. Use jsx for tree creation; edit for property updates on known nodes.",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "markup": {
+          "type": "string",
+          "description": "JSX-like nested markup string"
+        },
+        "parent": {
+          "type": "string",
+          "description": "Target parent node ID — the frame/container the new subtree will live inside"
+        },
+        "replaceId": {
+          "type": "string",
+          "description": "Replace this existing node in-place (keeps parent + sibling index). Old node is deleted on success. Markup must be single-root. Mutually exclusive with parent."
+        },
+        "insertIndex": {
+          "type": "number",
+          "description": "Position among parent siblings (0 = first). Omit to append at end. Ignored when replaceId is set (inherits old node index)."
+        }
+      },
+      "required": [
+        "markup"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "inspect",
+    "description": "Read design node(s) — choose what to surface with `facets`.\n\nDefault (no facets) returns a skeleton: id, name, type, role, children.\nFor anything else, list the facets you need — nothing else is included.\n\nFacets:\n  structure   name, type, size, layout shorthand — cheap overview\n  layout      layoutMode/gap/padding/align/sizing (row/column, fill/hug, etc.)\n  paint|fill  fills + Paint.boundVariables.color (see bound tokens)\n  stroke      strokes, strokeWeight, strokeAlign, dashPattern\n  effects     shadows, blurs\n  typography|text  fontFamily, fontSize, fontWeight, lineHeight, letterSpacing\n  appearance  opacity, visible, blendMode, cornerRadius, clipsContent\n  variables   node-level boundVariables + explicitVariableModes (token bindings)\n  all         everything\n\nParameters:\n  node    \"/\" for page root, or node ID from jsx/inspect results (e.g. \"100:5\").\n  facets  array of facet names listed above.\n  depth   Max tree depth (default: 5, max: 10).\n\nExamples:\n  inspect({node: \"/\"})                                   → page skeleton\n  inspect({node: \"100:5\"})                               → one-node skeleton\n  inspect({node: \"100:5\", facets: [\"variables\"]})        → token bindings only\n  inspect({node: \"100:5\", facets: [\"layout\", \"paint\"]})  → layout + fills\n  inspect({node: \"100:5\", facets: [\"all\"]})              → full properties\n\nUse `get_screenshot` for visual verification. Use `describe` for lint/validation.",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "\"/\" for page root, or node ID (e.g. \"100:5\")."
+        },
+        "facets": {
+          "type": "array",
+          "items": {
+            "type": "string",
+            "description": "A single facet name.",
+            "enum": [
+              "structure",
+              "paint",
+              "layout",
+              "text",
+              "effects",
+              "variables",
+              "appearance",
+              "stroke",
+              "fill",
+              "typography",
+              "all"
+            ]
+          },
+          "description": "Property buckets to surface. Omit for a skeleton-only response."
+        },
+        "depth": {
+          "type": "number",
+          "description": "Max depth (default: 5, max: 10)"
+        }
+      },
+      "required": [
+        "node"
+      ]
+    }
+  },
+  {
+    "name": "describe",
+    "description": "Validate a design subtree — semantic description, role detection, and lint rules. Checks for layout conflicts, overflow, missing properties, and structural issues.\n\nParameters:\n  node: Node ID to describe (e.g. \"100:5\"). Required.\n  depth: How deep to check children (default: 3, max: 8).\n\nReturns per-node: role, visual summary, layout summary, and issues (severity: error/warning/info).\n\nExamples:\n  describe({node: \"100:5\"})             → validate subtree, depth 3\n  describe({node: \"100:5\", depth: 1})   → shallow check (root + direct children only)",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID to describe (e.g. \"100:5\")."
+        },
+        "depth": {
+          "type": "number",
+          "description": "Max depth to check (default: 3, max: 8)"
+        }
+      },
+      "required": [
+        "node"
+      ]
+    }
+  },
+  {
+    "name": "edit",
+    "description": "Batch update properties on multiple nodes.\n\nFor single-property changes, prefer focused setters:\n  set_text  — text content\n  set_fill  — fill/background color\n  set_stroke — border\n  set_layout — padding, gap, direction\n\nUse edit for batch fixes or properties not covered by setters (sizing, radius, opacity, effects, component props):\n  edit({nodes: [\n    {node: \"1:1\", props: {w: \"fill\", corner: 8}},        // Figma native props\n    {node: \"1:2\", props: {opacity: 0.6}},\n    {node: \"1:3\", props: {Label: \"Sign In\"}},             // instance TEXT prop (by display name)\n  ]})\n\nFor instances, use component property DISPLAY NAMES (e.g. \"Label\") — edit resolves them to Figma's internal keys automatically. Component props can be mixed with Figma props in the same call.",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (e.g. \"1:2\") from jsx/inspect results"
+        },
+        "nodes": {
+          "type": "array",
+          "description": "Batch: array of {node, props?, content?} objects. No hard item cap — real ceiling is LLM output stream length (~10KB+ of rendered params can stall mid-JSON). If a batch is large and props are rich, split into 2 calls.",
+          "items": {
+            "type": "object",
+            "description": "{node, props?, content?} — at least one of props or content is required",
+            "properties": {
+              "node": {
+                "type": "string",
+                "description": "Node ID to update"
+              },
+              "props": {
+                "type": "object",
+                "description": "Properties to update (object, not a stringified JSON)"
+              },
+              "content": {
+                "type": "string",
+                "description": "New text content (for text nodes / overrides)"
+              }
+            },
+            "required": [
+              "node"
+            ]
+          }
+        },
+        "props": {
+          "type": "object",
+          "description": "Properties to update (single mode)"
+        },
+        "content": {
+          "type": "string",
+          "description": "New text content (single mode)"
+        }
+      }
+    },
+    "mutates": true
+  },
+  {
+    "name": "find_nodes",
+    "description": "Search nodes by name or type. Scoped to the current page — call switch_page first if your target lives on a different page.\n\nExamples:\n  find_nodes({query: \"Button\"})\n  find_nodes({query: \"frame\", scope: \"1:2\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "query": {
+          "type": "string",
+          "description": "Search query — matches node name or type"
+        },
+        "scope": {
+          "type": "string",
+          "description": "Limit search to subtree. Node ID (e.g. \"1:2\"). Default: entire page."
+        }
+      },
+      "required": [
+        "query"
+      ]
+    }
+  },
+  {
+    "name": "discover_props",
+    "description": "Discover unique property values in a subtree.\n\nExamples:\n  discover_props({node: \"1:2\", props: [\"fillColor\", \"fontSize\"]})\n\nSearchable properties: fillColor, textColor, strokeColor, strokeWeight, opacity, cornerRadius, gap, fontSize, fontFamily, fontWeight.",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Target node ID (e.g. \"1:2\")"
+        },
+        "props": {
+          "type": "array",
+          "description": "Properties to discover",
+          "items": {
+            "type": "string",
+            "description": "Property name"
+          }
+        }
+      },
+      "required": [
+        "node",
+        "props"
+      ]
+    }
+  },
+  {
+    "name": "replace_props",
+    "description": "Batch search-and-replace property values in a subtree.\n\nExamples:\n  replace_props({node: \"1:2\", rules: [{prop: \"fillColor\", from: \"#FFF\", to: \"#000\"}]})\n  replace_props({node: \"1:2\", rules: [\n    {prop: \"fillColor\", from: \"#FFF\", to: \"#000\"},\n    {prop: \"fontSize\", from: \"14\", to: \"16\"}\n  ]})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Target node ID (e.g. \"1:2\")"
+        },
+        "rules": {
+          "type": "array",
+          "description": "Replacement rules",
+          "items": {
+            "type": "object",
+            "description": "{prop, from, to}",
+            "properties": {
+              "prop": {
+                "type": "string",
+                "description": "Property name"
+              },
+              "from": {
+                "type": "string",
+                "description": "Value to find"
+              },
+              "to": {
+                "type": "string",
+                "description": "Value to replace with"
+              }
+            },
+            "required": [
+              "prop",
+              "from",
+              "to"
+            ]
+          }
+        }
+      },
+      "required": [
+        "node",
+        "rules"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "find_references",
+    "description": "Find every node on the current page that references a given variable.\n\nThis is the REVERSE of inspect: inspect asks \"what does this node bind?\",\nfind_references asks \"who uses this variable?\". Use it when renaming,\nauditing, or swapping tokens — you need to know all the binding sites\nbefore you touch the variable.\n\nScan scope: currentPage only. Invisible nodes are skipped by default.\nNode-level bindings (e.g. boundVariables.paddingLeft) and per-paint color\nbindings (fills[i].boundVariables.color, strokes[i].boundVariables.color)\nare both returned.\n\nParameters:\n  variable  — VariableID (e.g. \"VariableID:1:5\"). Required.\n\nReturns:\n  {variable, variableName, variableType, referenceCount,\n   references: [{nodeId, nodeName, nodeType, path}, ...]}\n\n  path values look like:\n    \"boundVariables.paddingLeft\"\n    \"fills[0].boundVariables.color\"\n    \"strokes[2].boundVariables.color\"\n\nExamples:\n  find_references({variable: \"VariableID:1:5\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "variable": {
+          "type": "string",
+          "description": "VariableID to look up (e.g. \"VariableID:1:5\"). Get IDs from list_variables."
+        }
+      },
+      "required": [
+        "variable"
+      ]
+    }
+  },
+  {
+    "name": "delete_node",
+    "description": "Delete a node and its children.\n\nExamples:\n  delete_node({node: \"1:2\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (e.g. \"1:2\")"
+        }
+      },
+      "required": [
+        "node"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "move_node",
+    "description": "Relocate a node without recreating it. Preserves IDs, bound variables, and component instances across the move, so callers tracking the node by ID never need to re-discover it. Use for: (a) changing child order within a container, (b) moving a subtree into a different parent, (c) fixing a placement mistake after jsx.\n\nExamples:\n  move_node({node: \"1:3\", name: \"NewTitle\"})         — rename in place\n  move_node({node: \"1:3\", parent: \"1:4\"})            — move into parent 1:4\n  move_node({node: \"1:5\", index: 0})                 — reorder within current parent",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (e.g. \"1:3\") to move/rename"
+        },
+        "parent": {
+          "type": "string",
+          "description": "Target parent node ID — the frame/container the node should live inside after the call"
+        },
+        "name": {
+          "type": "string",
+          "description": "New name (rename without changing parent)"
+        },
+        "index": {
+          "type": "number",
+          "description": "Reorder position among siblings. 0 = first, -1 = last."
+        }
+      },
+      "required": [
+        "node"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "clone_node",
+    "description": "Deep-copy a node with optional property overrides.\n\nExamples:\n  clone_node({node: \"1:2\"})                                 — clone to page root, same name\n  clone_node({node: \"1:2\", parent: \"/\"})                    — clone to page root explicitly\n  clone_node({node: \"1:2\", parent: \"/\", name: \"Hero Copy\"}) — clone to root with custom name\n  clone_node({node: \"1:2\", parent: \"1:4\"})                  — clone into parent node 1:4\n  clone_node({node: \"1:2\", parent: \"1:4\", overrides: {\"bg\": \"#D9D9D9\"}})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Source node ID (e.g. \"1:2\")"
+        },
+        "parent": {
+          "type": "string",
+          "description": "Target parent node ID the clone should live inside, or \"/\" for page root. Defaults to page root."
+        },
+        "name": {
+          "type": "string",
+          "description": "Name for the cloned node. Defaults to source node name."
+        },
+        "overrides": {
+          "type": "object",
+          "description": "Property overrides. Use \"Child.prop\" for child overrides."
+        }
+      },
+      "required": [
+        "node"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "skill",
+    "description": "Load a procedural skill — workflow + tool sequence + anti-patterns. Use FIRST when the user is changing/adjusting existing canvas, OR before creating a new design if a matching skill exists.\n\nExample: skill({ name: \"restyle\" })",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "name": {
+          "type": "string",
+          "description": "The skill name exactly as it appears in the KNOWLEDGE LIBRARY menu — no \"skill:\" prefix, no quotes."
+        }
+      },
+      "required": [
+        "name"
+      ]
+    }
+  },
+  {
+    "name": "style",
+    "description": "Load a visual style preset — color tokens, typography, shape, depth. Use when picking a named aesthetic from the menu before generating new design.\n\nExample: style({ name: \"neon-cyber\" })",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "name": {
+          "type": "string",
+          "description": "The style name exactly as it appears in the KNOWLEDGE LIBRARY menu — no \"style:\" prefix, no quotes."
+        }
+      },
+      "required": [
+        "name"
+      ]
+    }
+  },
+  {
+    "name": "anatomy",
+    "description": "Load a component anatomy reference — structural blueprint of a UI component (parts, slots, hierarchy). Use before building complex components.\n\nExample: anatomy({ name: \"data-table\" })",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "name": {
+          "type": "string",
+          "description": "The anatomy name exactly as it appears in the KNOWLEDGE LIBRARY menu — no \"anatomy:\" prefix, no quotes."
+        }
+      },
+      "required": [
+        "name"
+      ]
+    }
+  },
+  {
+    "name": "guideline",
+    "description": "Load a page-type design guideline — layout patterns for landing pages, dashboards, login flows, forms, etc.\n\nExample: guideline({ name: \"form\" })",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "name": {
+          "type": "string",
+          "description": "The guideline name exactly as it appears in the KNOWLEDGE LIBRARY menu — no \"guideline:\" prefix, no quotes."
+        }
+      },
+      "required": [
+        "name"
+      ]
+    }
+  },
+  {
+    "name": "help",
+    "description": "Load narrow how-to / process help — tool usage rules, edge cases, naming conventions.\n\nExample: help({ name: \"interaction-model\" })",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "name": {
+          "type": "string",
+          "description": "The help name exactly as it appears in the KNOWLEDGE LIBRARY menu — no \"help:\" prefix, no quotes."
+        }
+      },
+      "required": [
+        "name"
+      ]
+    }
+  },
+  {
+    "name": "list_variables",
+    "description": "List variables as a flat array with referenced collections.\n\nReturns {data: {variables[], collections[], nextCursor?}}. Each variable carries\nits full Figma shape: id, name, variableCollectionId, resolvedType, valuesByMode.\ncollections[] only includes collections referenced by the returned variables\n(use for mode-name resolution).\n\nParameters:\n  collection — VariableCollectionId to filter by\n  filter     — substring match on variable name (case-insensitive)\n  cursor     — opaque pagination cursor from a previous call\n  limit      — max variables per page (default 100)\n\nExamples:\n  list_variables()\n  list_variables({collection: \"VariableCollectionId:1:2\"})\n  list_variables({filter: \"bg\"})\n  list_variables({cursor: \"100\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "collection": {
+          "type": "string",
+          "description": "VariableCollectionId to filter by"
+        },
+        "filter": {
+          "type": "string",
+          "description": "Substring match on variable name (case-insensitive)"
+        },
+        "cursor": {
+          "type": "string",
+          "description": "Opaque pagination cursor from a previous call"
+        },
+        "limit": {
+          "type": "number",
+          "description": "Max variables per page (default 100)"
+        }
+      }
+    }
+  },
+  {
+    "name": "create_collection",
+    "description": "Create a VariableCollection with named modes.\n\nThe first mode in the array becomes the default mode. Returns\n{data: {id, modes: [{modeId, name}]}} — use those modeIds with\nset_variable_value and set_variable_mode.\n\nExamples:\n  create_collection({name: \"Theme\", modes: [\"Light\", \"Dark\"]})\n  create_collection({name: \"Device\", modes: [\"Desktop\", \"Tablet\", \"Mobile\"]})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "name": {
+          "type": "string",
+          "description": "Collection name"
+        },
+        "modes": {
+          "type": "array",
+          "description": "Mode names (first becomes default)",
+          "items": {
+            "type": "string",
+            "description": "Mode name"
+          }
+        }
+      },
+      "required": [
+        "name",
+        "modes"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "create_variable",
+    "description": "Create a variable in an existing collection.\n\nNo value is set here — use set_variable_value after. Returns {data: {id}}.\n\nExamples:\n  create_variable({collection: \"VariableCollectionId:1:2\", name: \"Theme/bg\", type: \"COLOR\"})\n  create_variable({collection: \"VariableCollectionId:1:2\", name: \"spacing/md\", type: \"FLOAT\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "collection": {
+          "type": "string",
+          "description": "VariableCollectionId to create the variable in"
+        },
+        "name": {
+          "type": "string",
+          "description": "Variable name (slashes denote hierarchy in the Figma UI)"
+        },
+        "type": {
+          "type": "string",
+          "description": "Variable type",
+          "enum": [
+            "COLOR",
+            "FLOAT",
+            "BOOLEAN",
+            "STRING"
+          ]
+        }
+      },
+      "required": [
+        "collection",
+        "name",
+        "type"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "ensure_collection",
+    "description": "Idempotent VariableCollection creation — safe to retry.\n\nReturns existing collection if one with the same name + identical mode list\nalready exists, otherwise creates a new one. Spec §3.1.\n\nPrefer this over create_collection — re-running with the same name + modes\nreturns the existing collection instead of creating a duplicate.\n\nOmit idempotency_key — the handler computes it canonically from (name, modes).\nPass it only if you need strict concurrency-safety validation (LLMs should\nnot try to compute SHA-256 inline; placeholder strings are rejected).\n\nReturns {data: {collection_id, modes: [{modeId, name}], reused?: true}}.\n\nExamples:\n  ensure_collection({name: \"Theme\", modes: [\"Light\", \"Dark\"]})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "name": {
+          "type": "string",
+          "description": "Collection name"
+        },
+        "modes": {
+          "type": "array",
+          "description": "Mode names (first becomes default)",
+          "items": {
+            "type": "string",
+            "description": "Mode name"
+          }
+        },
+        "idempotency_key": {
+          "type": "string",
+          "description": "Optional. Handler auto-computes canonically if omitted. Pass only if you need strict concurrency-safety validation against the SHA-256 formula in spec §3.1."
+        }
+      },
+      "required": [
+        "name",
+        "modes"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "ensure_variable",
+    "description": "Idempotent variable creation — populates values_by_mode in one shot.\n\nPrefer this over create_variable — re-running with the same args returns the\nexisting variable instead of creating a duplicate.\n\nBehavior (spec §3.1):\n  - Exactly 1 variable with (collection_id, name, type) in target collection → idempotent reuse.\n  - 0 in target, matches in OTHER collections → create new in target + warning NAME_EXISTS_OUTSIDE_TARGET_COLLECTION.\n  - 0 anywhere → create new.\n  - 2+ in target collection (Figma allows duplicates) → fail SAME_COLLECTION_NAME_DUPLICATE.\n\nvalues_by_mode keys can be either mode NAMES (e.g. \"Light\") or modeIds (e.g. \"1:0\").\nEach value must match the variable type (hex/RGBA for COLOR, number for FLOAT,\nstring for STRING, boolean for BOOLEAN).\n\nOmit idempotency_key — the handler computes it canonically from\n(collection_id, name, type, values_by_mode). Pass it only if you need strict\nconcurrency-safety validation (LLMs should not try to compute SHA-256 inline;\nplaceholder strings are rejected).\n\nMode coverage policy (spec §6.2):\n  - mode_coverage_required: 'all' (default) — every mode in the collection\n    must have an explicit value. set_fill / bind_variable will REJECT\n    bindings that fall through to a missing mode (MISSING_MODE_VALUES).\n  - mode_coverage_required: 'opt-in-fallback' — fallback to default mode\n    is intended. Bindings emit FALLBACK_BINDING warning instead of failing.\n    Caller MUST provide fallback_reason containing the structured phrase\n    \"fallback to <mode_name>\" (machine-greppable).\n\nReturns {data: {variable_id, name, type, collection_id, mode_coverage[],\nmode_coverage_required, reused?: true}, warnings?: [...]}.\n\nExamples:\n  ensure_variable({collection_id: \"VariableCollectionId:1:2\", name: \"Text/Primary\", type: \"COLOR\", values_by_mode: {Light: \"#111\", Dark: \"#EEE\"}})\n  ensure_variable({collection_id: \"VariableCollectionId:1:2\", name: \"Spacing/desktop\", type: \"FLOAT\", values_by_mode: {Desktop: 24}, mode_coverage_required: \"opt-in-fallback\", fallback_reason: \"Desktop-only metric; fallback to Desktop in Mobile mode.\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "collection_id": {
+          "type": "string",
+          "description": "Target VariableCollectionId — strict ID, no name lookup."
+        },
+        "name": {
+          "type": "string",
+          "description": "Variable name (slashes denote hierarchy)."
+        },
+        "type": {
+          "type": "string",
+          "description": "Variable type",
+          "enum": [
+            "COLOR",
+            "FLOAT",
+            "STRING",
+            "BOOLEAN"
+          ]
+        },
+        "values_by_mode": {
+          "type": "object",
+          "description": "Map of mode name OR modeId → value. Hex strings allowed for COLOR."
+        },
+        "idempotency_key": {
+          "type": "string",
+          "description": "Optional. Handler auto-computes canonically if omitted. Pass only if you need strict concurrency-safety validation against the SHA-256 formula in spec §3.1."
+        },
+        "mode_coverage_required": {
+          "type": "string",
+          "description": "Mode coverage policy — \"all\" (default; every mode must have explicit value) or \"opt-in-fallback\" (allow fallback, requires fallback_reason).",
+          "enum": [
+            "all",
+            "opt-in-fallback"
+          ]
+        },
+        "fallback_reason": {
+          "type": "string",
+          "description": "REQUIRED iff mode_coverage_required=\"opt-in-fallback\". Must contain the structured phrase \"fallback to <mode_name>\". Persisted on the variable for audit trail."
+        }
+      },
+      "required": [
+        "collection_id",
+        "name",
+        "type",
+        "values_by_mode"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "set_variable_value",
+    "description": "Set a variable's value for a specific mode.\n\nThin wrapper over Figma's variable.setValueForMode(modeId, value). Call once per\nmode. Value is a raw value (COLOR/FLOAT/STRING/BOOLEAN) OR an alias object\n{type: \"VARIABLE_ALIAS\", id: \"VariableID:x:y\"}. Hex strings are accepted for\nCOLOR and normalized to {r,g,b,a} in 0-1 range.\n\nExamples:\n  set_variable_value({variable: \"VariableID:1:5\", mode: \"1:0\", value: \"#FFFFFF\"})\n  set_variable_value({variable: \"VariableID:1:5\", mode: \"1:1\", value: {r: 0.1, g: 0.1, b: 0.1, a: 1}})\n  set_variable_value({variable: \"VariableID:1:6\", mode: \"1:0\", value: 16})\n  set_variable_value({variable: \"VariableID:1:7\", mode: \"1:0\", value: {type: \"VARIABLE_ALIAS\", id: \"VariableID:1:9\"}})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "variable": {
+          "type": "string",
+          "description": "VariableID to set"
+        },
+        "mode": {
+          "type": "string",
+          "description": "Mode id from the variable's collection (e.g. \"1:0\")"
+        },
+        "value": {
+          "type": "object",
+          "description": "Raw value (COLOR/FLOAT/STRING/BOOLEAN) or {type: \"VARIABLE_ALIAS\", id}"
+        }
+      },
+      "required": [
+        "variable",
+        "mode",
+        "value"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "bind_variable",
+    "description": "Bind a FLOAT, BOOLEAN, or STRING variable to a node property.\n\nprop is a flat Figma bindable field (e.g. fontSize, itemSpacing, paddingTop,\ncornerRadius, opacity, visible, width, height, characters). Shorthands:\ngap → itemSpacing, padding → paddingTop, corner → cornerRadius,\nfont-size → fontSize.\n\nCOLOR variables are NOT bound here — they live inside Paint objects. To apply\na color token, specify it at the source instead:\n  • At creation:  jsx <frame bg=\"$TokenName\" ...> or fill=\"$TokenName\"\n  • Post-hoc:     set_fill({node, bg: \"$TokenName\"}) or set_stroke\n\nWhen selecting which variable to bind: if the node is a Tablet or Mobile variant\n(name or variant property contains \"Tablet\"/\"Mobile\"), match the node's property\nvalue against the Tablet/Mobile mode column from list_variables — not Desktop.\n\nExamples:\n  bind_variable({node: \"1:2\", prop: \"fontSize\", variable: \"VariableID:1:6\"})\n  bind_variable({node: \"1:3\", prop: \"paddingTop\", variable: \"VariableID:1:7\"})\n  bind_variable({node: \"1:4\", prop: \"visible\", variable: \"VariableID:1:8\"})\n  bind_variable({node: \"1:5\", prop: \"characters\", variable: \"VariableID:1:9\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (e.g. \"1:2\")"
+        },
+        "prop": {
+          "type": "string",
+          "description": "Flat Figma bindable field (fontSize, paddingTop, itemSpacing, visible, characters, etc.). COLOR props (fills/strokes) not supported — use set_fill/jsx."
+        },
+        "variable": {
+          "type": "string",
+          "description": "VariableID to bind (FLOAT/BOOLEAN/STRING only)"
+        }
+      },
+      "required": [
+        "node",
+        "prop",
+        "variable"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "set_variable_mode",
+    "description": "Set a node to use a specific mode of a variable collection.\n\nThis controls which variable values the node displays. For example, set a frame\nto use \"Dark\" mode of the \"Theme\" collection so all bound variables show dark values.\n\nExamples:\n  set_variable_mode({node: \"1:2\", collection: \"VariableCollectionId:1:2\", mode: \"1:1\"})\n  set_variable_mode({node: \"1:5\", collection: \"VariableCollectionId:1:3\", mode: \"1:2\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (e.g. \"1:2\")"
+        },
+        "collection": {
+          "type": "string",
+          "description": "VariableCollectionId"
+        },
+        "mode": {
+          "type": "string",
+          "description": "Mode id (e.g. \"1:1\")"
+        }
+      },
+      "required": [
+        "node",
+        "collection",
+        "mode"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "create_component",
+    "description": "Convert a frame or group to a Figma component.\n\nExamples:\n  create_component({node: \"1:2\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (e.g. \"1:2\") to convert"
+        }
+      },
+      "required": [
+        "node"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "combine_components",
+    "description": "Combine multiple components into a variant set (ComponentSet).\n\nExamples:\n  combine_components({nodes: [\"1:2\", \"1:3\", \"1:4\"], name: \"Button\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "nodes": {
+          "type": "array",
+          "description": "Component node IDs to combine",
+          "items": {
+            "type": "string",
+            "description": "Node ID"
+          }
+        },
+        "name": {
+          "type": "string",
+          "description": "Component set name"
+        }
+      },
+      "required": [
+        "nodes"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "add_component_prop",
+    "description": "Add a component property and bind it to a child node.\n\nFor TEXT properties: binds to the target text node's characters, so instances can override the text content.\nFor BOOLEAN properties: binds to the target node's visibility.\n\nParameters:\n  node: The component node ID (must be COMPONENT or COMPONENT_SET).\n  name: Property display name.\n  type: TEXT, BOOLEAN, or INSTANCE_SWAP.\n  default: Default value.\n  bind: Child node ID to bind this property to. For TEXT, binds to text content. For BOOLEAN, binds to visibility.\n\nExamples:\n  add_component_prop({node: \"1:2\", name: \"Label\", type: \"TEXT\", default: \"Click me\", bind: \"1:5\"})\n  add_component_prop({node: \"1:2\", name: \"Show Icon\", type: \"BOOLEAN\", default: \"true\", bind: \"1:6\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Component node ID (e.g. \"1:2\")"
+        },
+        "name": {
+          "type": "string",
+          "description": "Property name"
+        },
+        "type": {
+          "type": "string",
+          "description": "Property type",
+          "enum": [
+            "TEXT",
+            "BOOLEAN",
+            "INSTANCE_SWAP"
+          ]
+        },
+        "default": {
+          "type": "string",
+          "description": "Default value"
+        },
+        "bind": {
+          "type": "string",
+          "description": "Child node ID to bind this property to"
+        }
+      },
+      "required": [
+        "node",
+        "name",
+        "type"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "list_component_props",
+    "description": "List properties and variants of a component, component set, or instance.\n\nExamples:\n  list_component_props({node: \"1:2\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Component/instance node ID (e.g. \"1:2\")"
+        }
+      },
+      "required": [
+        "node"
+      ]
+    }
+  },
+  {
+    "name": "create_instance",
+    "description": "Create an instance of a component.\n\nExamples:\n  create_instance({node: \"1:2\"})\n  create_instance({node: \"1:2\", parent: \"1:4\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Component node ID (e.g. \"1:2\") to instantiate"
+        },
+        "parent": {
+          "type": "string",
+          "description": "Parent node ID for placement"
+        }
+      },
+      "required": [
+        "node"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "set_text",
+    "description": "Set text content on one or more nodes.\n\n  set_text({node: \"1:2\", text: \"Hello World\"})\n  set_text({nodes: [{node: \"1:2\", text: \"Title\"}, {node: \"1:3\", text: \"Subtitle\"}]})\n\nUse this when changing what text says. For text styling (font, size, weight), use edit.",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID"
+        },
+        "text": {
+          "type": "string",
+          "description": "New text content"
+        },
+        "nodes": {
+          "type": "array",
+          "description": "Batch: [{node, text}]",
+          "items": {
+            "type": "object",
+            "description": "{node, text}"
+          }
+        }
+      }
+    },
+    "mutates": true
+  },
+  {
+    "name": "set_fill",
+    "description": "Set fill or background color on a node.\n\n  set_fill({node: \"1:2\", bg: \"#F5F5F5\"})\n  set_fill({node: \"1:2\", fill: \"#333333\"})\n  set_fill({node: \"1:2\", bg: \"linear-gradient(135deg, #8B5CF6 0%, #F97316 100%)\"})\n\n  // Batch — bulk paint update in one call:\n  set_fill({nodes: [{node: \"1:2\", bg: \"#FFF\"}, {node: \"1:3\", bg: \"#F5F5F5\"}]})\n\nfill = text color or shape fill. bg = frame background.\nFor stroke color, use set_stroke.\n\nAccepted color formats (for fill or bg):\n  hex                \"#FFF\", \"#F5F5F5\"\n  gradient string    \"linear-gradient(angle, #color stop%, ...)\" or \"radial-gradient(...)\"\n  variable token     qualified bare name \"$Surface/Card\"\n  transparent        \"transparent\" (bg only)",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (single mode)"
+        },
+        "fill": {
+          "type": "string",
+          "description": "Text color or shape fill — hex, gradient string, or qualified bare-name token"
+        },
+        "bg": {
+          "type": "string",
+          "description": "Background — hex, gradient string, \"transparent\", or qualified bare-name token"
+        },
+        "nodes": {
+          "type": "array",
+          "description": "Batch: [{node, fill?, bg?}]",
+          "items": {
+            "type": "object",
+            "description": "{node, fill?, bg?}"
+          }
+        }
+      }
+    },
+    "mutates": true
+  },
+  {
+    "name": "set_stroke",
+    "description": "Set stroke (border) on a node.\n\n  set_stroke({node: \"1:2\", stroke: \"1 #E0E0E0\"})\n  set_stroke({node: \"1:2\", stroke: \"2 #333 inside\"})\n  set_stroke({node: \"1:2\", color: \"#E0E0E0\", weight: 1, align: \"inside\"})\n  set_stroke({node: \"1:2\", color: \"linear-gradient(90deg, #8B5CF6 0%, #F97316 100%)\", weight: 1.5, align: \"inside\"})\n\n  // Batch — bulk stroke update in one call:\n  set_stroke({nodes: [{node: \"1:2\", color: \"#E0E0E0\", weight: 1}, {node: \"1:3\", color: \"#333\", weight: 2}]})\n\nShorthand: \"weight color align\" (e.g. \"1 #E0E0E0 inside\"). Hex only in shorthand.\n\nAccepted color formats (for the explicit `color` field, not the shorthand):\n  hex                \"#E0E0E0\"\n  gradient string    \"linear-gradient(angle, #color stop%, ...)\" or \"radial-gradient(...)\"\n  variable token     qualified bare name \"$Border/Default\"\n\nTo bind a variable to the stroke color, use the explicit `color` field — the shorthand parser silently drops bare-name tokens.",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (single mode)"
+        },
+        "stroke": {
+          "type": "string",
+          "description": "Shorthand: \"1 #E0E0E0 inside\" — single-string form. Hex only; for gradient/variable use the explicit color field."
+        },
+        "color": {
+          "type": "string",
+          "description": "Stroke color — hex, gradient string, or qualified bare-name token"
+        },
+        "weight": {
+          "type": "number",
+          "description": "Stroke weight in px"
+        },
+        "align": {
+          "type": "string",
+          "enum": [
+            "inside",
+            "outside",
+            "center"
+          ],
+          "description": "Stroke alignment relative to the frame edge"
+        },
+        "nodes": {
+          "type": "array",
+          "description": "Batch: [{node, color?, weight?, align?, stroke?}]",
+          "items": {
+            "type": "object",
+            "description": "{node, color?, weight?, align?, stroke?}"
+          }
+        }
+      }
+    },
+    "mutates": true
+  },
+  {
+    "name": "set_layout",
+    "description": "Set auto-layout properties on a container.\n\n  set_layout({node: \"1:2\", gap: 16, p: 24})\n  set_layout({node: \"1:2\", layout: \"row\", justify: \"space-between\"})\n  set_layout({node: \"1:2\", layout: \"column\", gap: 8, p: \"16 24\", align: \"center\"})\n  set_layout({node: \"1:2\", layout: \"grid\", cols: 3, rows: 2, gap: 16})\n\n  // Batch — bulk update in one call:\n  set_layout({nodes: [{node: \"1:2\", gap: 16, p: 24}, {node: \"1:3\", gap: 8, p: 12}]})\n\nControls spacing, padding, direction, and alignment of a container's children.\nGrid: use layout:\"grid\" with cols/rows + gap (or rowGap/colGap for asymmetric).\nChildren fill the grid in insertion order.",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (single mode)"
+        },
+        "layout": {
+          "type": "string",
+          "enum": [
+            "row",
+            "column",
+            "grid"
+          ],
+          "description": "Auto-layout mode"
+        },
+        "gap": {
+          "type": "number",
+          "description": "Spacing between children (px). On grid sets both row+column gap."
+        },
+        "rowGap": {
+          "type": "number",
+          "description": "Grid row gap (px, grid only)"
+        },
+        "colGap": {
+          "type": "number",
+          "description": "Grid column gap (px, grid only)"
+        },
+        "cols": {
+          "type": "number",
+          "description": "Grid column count (required when layout=\"grid\")"
+        },
+        "rows": {
+          "type": "number",
+          "description": "Grid row count (required when layout=\"grid\")"
+        },
+        "p": {
+          "type": "number",
+          "description": "Padding — number, \"v h\", or \"t r b l\""
+        },
+        "justify": {
+          "type": "string",
+          "enum": [
+            "center",
+            "space-between",
+            "start",
+            "end"
+          ],
+          "description": "Main axis (flex only)"
+        },
+        "align": {
+          "type": "string",
+          "enum": [
+            "center",
+            "start",
+            "end",
+            "baseline"
+          ],
+          "description": "Cross axis (flex only)"
+        },
+        "wrap": {
+          "type": "string",
+          "enum": [
+            "wrap",
+            "nowrap"
+          ],
+          "description": "Wrap behaviour (flex only)"
+        },
+        "nodes": {
+          "type": "array",
+          "description": "Batch: [{node, layout?, gap?, p?, ...}]",
+          "items": {
+            "type": "object",
+            "description": "{node, layout?, gap?, p?, ...}"
+          }
+        }
+      }
+    },
+    "mutates": true
+  },
+  {
+    "name": "get_selection",
+    "description": "Get the user's currently selected nodes in Figma.\n\nReturns node names, types, and IDs of selected elements.\nCall this when the user's intent involves modifying existing elements:\n- \"change this button\", \"update the card\", \"fix the spacing\"\n- References to \"this\", \"the selected\", \"it\"\n\nSkip for fresh design requests (\"design a login page\", \"create a dashboard\") — a new canvas has no selection to read, so the call returns nothing and burns an iteration.\n\nExamples:\n  get_selection()",
+    "parameters": {
+      "type": "object",
+      "properties": {}
+    }
+  },
+  {
+    "name": "switch_page",
+    "description": "Navigate between pages in the Figma file. ID-driven — names are not addressable (they can collide and change).\n\nTwo modes:\n- switch_page({})              → return the page roster only, no switch (use to discover IDs on first call)\n- switch_page({pageId: \"1:23\"})  → switch and return the updated state + roster\n\nPages are top-level containers under the file root. Most read/write operations default to figma.currentPage. Call this when you need to operate on nodes that live on a different page than the current one.\n\nReturns:\n- currentPageId, currentPageName — the now-current page (always present)\n- pages — full roster [{id, name}] of every page in the file (always present)\n- previousPageId, previousPageName — what you switched from (only when an actual switch happened)\n- unchanged — true if target was already current\n\nTypical flow:\n1. switch_page({})                         // get IDs\n2. switch_page({pageId: \"<picked id>\"})    // switch\n\nWhen to call:\n- User mentions content on a different page than the current one\n- A previous tool reported a node ID is on a non-current page\n- You need to inspect/modify nodes outside the active page\n\nDon't call:\n- For nodes already on the current page — figma.currentPage is the default scope, this would just waste an iteration\n- Repeatedly to \"explore\" — every call returns the full pages roster, cache it",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "pageId": {
+          "type": "string",
+          "description": "Target page ID (e.g., \"0:1\"). Omit to just fetch the page roster without switching."
+        }
+      }
+    },
+    "mutates": false
+  },
+  {
+    "name": "read_plugin_data",
+    "description": "Read plugin data (private or shared) from a Figma node.\n\nUse for: i18n metadata, design-system tags, custom plugin annotations, anything stored via setPluginData / setSharedPluginData.\n\nIf `namespace` is omitted, reads private pluginData (`node.getPluginData(key)`).\nIf `namespace` is provided, reads sharedPluginData (`node.getSharedPluginData(namespace, key)`).\n\nReturns `{value: \"\"}` (empty string) when the key does not exist — Figma's API never throws here.\n\nExamples:\n  read_plugin_data({node_id: \"1:5\", key: \"ref\"})\n  read_plugin_data({node_id: \"1:5\", namespace: \"i18n\", key: \"ref\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node_id": {
+          "type": "string",
+          "description": "Figma node id (resolve via find_nodes / get_selection first)."
+        },
+        "namespace": {
+          "type": "string",
+          "description": "Optional sharedPluginData namespace. Omit for private pluginData."
+        },
+        "key": {
+          "type": "string",
+          "description": "Key to read."
+        }
+      },
+      "required": [
+        "node_id",
+        "key"
+      ]
+    },
+    "mutates": false
+  },
+  {
+    "name": "write_plugin_data",
+    "description": "Write plugin data (private or shared) to a Figma node.\n\nIf `namespace` is omitted, writes private pluginData (`node.setPluginData(key, value)`).\nIf `namespace` is provided, writes sharedPluginData (`node.setSharedPluginData(namespace, key, value)`).\n\nPass an empty string as `value` to delete a key.\n\nExamples:\n  write_plugin_data({node_id: \"1:5\", key: \"ref\", value: \"home_title\"})\n  write_plugin_data({node_id: \"1:5\", namespace: \"i18n\", key: \"ref\", value: \"home.welcome_title\"})",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node_id": {
+          "type": "string",
+          "description": "Figma node id."
+        },
+        "namespace": {
+          "type": "string",
+          "description": "Optional sharedPluginData namespace."
+        },
+        "key": {
+          "type": "string",
+          "description": "Key to write."
+        },
+        "value": {
+          "type": "string",
+          "description": "Value (string). Pass \"\" to delete."
+        }
+      },
+      "required": [
+        "node_id",
+        "key",
+        "value"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "create_vector",
+    "description": "Create a vector node from SVG path data or a list of points. Use for chart lines, custom icon paths, freeform curves, or any shape that needs path data.\n\nExamples:\n  // Polyline (chart trend line)\n  create_vector({\n    parent: \"1:23\", name: \"TrendLine\",\n    x: 40, y: 20, width: 550, height: 240,\n    points: [[0,144],[90,96],[180,120],[270,64],[360,80],[450,40],[540,72]],\n    stroke: \"#6366F1\", strokeWeight: 2\n  })\n\n  // Raw SVG path (custom shape)\n  create_vector({\n    parent: \"1:23\", name: \"Wave\",\n    width: 200, height: 60,\n    data: \"M 0 30 Q 50 0 100 30 T 200 30\",\n    stroke: \"linear-gradient(90deg, #8B5CF6 0%, #F97316 100%)\",\n    strokeWeight: 1.5\n  })\n\nPath input — provide ONE of:\n  points: [[x,y], ...]   compiled to \"M x0 y0 L x1 y1 ...\" (polyline shortcut)\n  data:   \"M ... L ...\"  raw SVG path (LLM-native; supports M, L, C, Q, A, Z)\n\nStroke / fill (same formats as set_stroke / set_fill):\n  hex            \"#6366F1\"\n  gradient       \"linear-gradient(angle, #color stop%, ...)\"\n  variable       qualified bare name \"$Brand/Primary\"\n\nDefault fill is \"transparent\" so the vector shows only its stroke. Pass an explicit fill if you want it filled.\n\nWhen NOT to use:\n  - Standard rectangles / ellipses / lines — use jsx <Rect/>, <Ellipse/>, <Line/> elements (simpler, batch-friendly)\n  - Existing vector edits — use edit / set_stroke instead",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "parent": {
+          "type": "string",
+          "description": "Parent node ID. Omit to attach to the current page."
+        },
+        "name": {
+          "type": "string",
+          "description": "Node name (default: \"Vector\")."
+        },
+        "x": {
+          "type": "number",
+          "description": "X position relative to parent (default: 0)."
+        },
+        "y": {
+          "type": "number",
+          "description": "Y position relative to parent (default: 0)."
+        },
+        "width": {
+          "type": "number",
+          "description": "Vector bounds width in px."
+        },
+        "height": {
+          "type": "number",
+          "description": "Vector bounds height in px."
+        },
+        "data": {
+          "type": "string",
+          "description": "Raw SVG path string. Mutually exclusive with `points`."
+        },
+        "points": {
+          "type": "array",
+          "description": "Polyline points as [[x,y], ...]. Compiled internally to \"M x0 y0 L x1 y1 ...\". Mutually exclusive with `data`.",
+          "items": {
+            "type": "array",
+            "description": "[x, y] pair",
+            "items": {
+              "type": "number",
+              "description": "coordinate"
+            }
+          }
+        },
+        "windingRule": {
+          "type": "string",
+          "enum": [
+            "NONZERO",
+            "EVENODD"
+          ],
+          "description": "Path fill winding rule (default: NONZERO)."
+        },
+        "stroke": {
+          "type": "string",
+          "description": "Stroke color — hex, gradient string, or qualified bare-name token."
+        },
+        "strokeWeight": {
+          "type": "number",
+          "description": "Stroke weight in px (default: 1)."
+        },
+        "strokeAlign": {
+          "type": "string",
+          "enum": [
+            "inside",
+            "outside",
+            "center"
+          ],
+          "description": "Stroke alignment relative to the path (default: center)."
+        },
+        "fill": {
+          "type": "string",
+          "description": "Fill — hex / gradient / variable / \"transparent\" (default: \"transparent\")."
+        }
+      },
+      "required": [
+        "width",
+        "height"
+      ]
+    },
+    "mutates": true
+  },
+  {
+    "name": "get_screenshot",
+    "description": "Capture a PNG screenshot of a node.\n\nUse after style changes to visually verify the result instead of reading properties back.\nReturns base64 PNG data embedded in the response.\n\nParameters:\n  node: Node ID from jsx/inspect results (e.g. \"100:5\"). Page root (\"/\") is not supported.\n  scale: Export scale 0.5–2 (default 1). Higher = larger file.\n  padding: Reserved for future use — currently ignored.\n\nExamples:\n  get_screenshot({node: \"100:5\"})              → PNG at 1x\n  get_screenshot({node: \"100:5\", scale: 2})    → PNG at 2x (sharper)",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "node": {
+          "type": "string",
+          "description": "Node ID (e.g. \"100:5\"). Page root \"/\" is not supported."
+        },
+        "scale": {
+          "type": "number",
+          "description": "Export scale 0.5–2 (default 1)."
+        },
+        "padding": {
+          "type": "number",
+          "description": "Reserved for future use."
+        }
+      },
+      "required": [
+        "node"
+      ]
+    },
+    "mutates": false
+  },
+  {
+    "name": "ask_user",
+    "description": "Ask the user 1-3 questions in a single form. Each question has its own options and can be single- or multi-select. Bundle related decisions in ONE call instead of multiple turns.\n\nUse when:\n- The prompt is ambiguous on multiple dimensions (audience + aesthetic + length) — bundle them into one form\n- You need a decision before proceeding (delete existing? which section first?)\n- Multiple valid approaches exist and user preference matters\n\nReturns one of:\n- { answers: [...] } — array indexed to questions order. string for single-select, string[] for multi-select. The string MAY be one of the option labels OR custom text the user typed via the auto-injected \"Other...\" option.\n- { freeText: \"...\" } — when the user typed a free-form answer in the chat input instead of submitting the form. Treat as authoritative — user is overriding the structured options.\n\nEach question:\n- question: required prompt string. Self-contained — no separate header/label, the question text IS the heading.\n- options: 2-3 options, each { label, description? }. The form auto-injects an \"Other...\" row, so the user always sees options.length + 1 rows total — keep options ≤ 3 to stay within the 4-row visual cap.\n- multiSelect: optional boolean (default false). Use only when the answer is genuinely a list (e.g. \"which features?\"). For mutually exclusive choices keep false.\n\nConventions:\n- **First option = recommended.** If you have a strong default for the user, put it FIRST and add \"(Recommended)\" at the end of the label. The form auto-focuses the first option and the dev/auto-pick fallback selects it — both work better with a deliberate recommendation.\n- **Do NOT include an \"Other\" option yourself** — the form auto-injects an \"Other...\" row per question with an inline text input. Don't add a redundant one.\n- **Bundle aggressively.** 2 related dimensions in ONE call beats 2 sequential turns.\n\nExample:\n  ask_user({ questions: [\n    { question: \"Who is this for?\", options: [{label:\"B2B SaaS\"},{label:\"Consumer\"},{label:\"Developer tool\"}] },\n    { question: \"What visual direction?\", options: [{label:\"Minimal\"},{label:\"Bold/Brutalist\"},{label:\"Neon/Cyber\"}] }\n  ]})\n\nSkip when the prompt is already actionable — asking adds a turn and costs momentum.",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "questions": {
+          "type": "array",
+          "description": "1-3 questions to present in a single form (3 is the soft cap; bundle related decisions but don't pad).",
+          "maxItems": 3,
+          "items": {
+            "type": "object",
+            "description": "A single question with its own options and select mode",
+            "properties": {
+              "question": {
+                "type": "string",
+                "description": "The question prompt"
+              },
+              "options": {
+                "type": "array",
+                "description": "2-3 options for this question. The form auto-injects an \"Other...\" row, so user-visible total is options.length + 1 (cap at 4).",
+                "maxItems": 3,
+                "items": {
+                  "type": "object",
+                  "description": "A single option (label + optional description)",
+                  "properties": {
+                    "label": {
+                      "type": "string",
+                      "description": "Short option label, ~6-20 chars. The label IS the heading — no need for a separate header."
+                    },
+                    "description": {
+                      "type": "string",
+                      "description": "Brief one-line explanation, ~50-80 chars max. Anything longer is clamped to 2 lines with ellipsis in the UI."
+                    }
+                  },
+                  "required": [
+                    "label"
+                  ]
+                }
+              },
+              "multiSelect": {
+                "type": "boolean",
+                "description": "When true, allow multiple options to be selected. Default false (single-select)."
+              }
+            },
+            "required": [
+              "question",
+              "options"
+            ]
+          }
+        }
+      },
+      "required": [
+        "questions"
+      ]
+    },
+    "mutates": false
+  },
+  {
+    "name": "subtask",
+    "description": "Delegate a focused sub-task to a typed child agent. Each type has its own tools, iteration budget, and behavioral constraints.\n\nAvailable agent types:\n- create: Build an independent UI section (header, sidebar, form, card). Default.\n- audit: Read-only design review — find layout issues, property omissions, report PASS/FAIL.\n- token: Variable system operations — create collections, bind tokens, set up aliases.\n\nUse when the prompt names 3+ distinct regions (e.g. header, sidebar, main) that share no nodes, or when specialized behavior is needed (audit, token ops). For 1-2 tool-call operations, inline calls finish faster than the subtask spin-up cost.",
+    "parameters": {
+      "type": "object",
+      "properties": {
+        "prompt": {
+          "type": "string",
+          "description": "Description of the sub-task to delegate. Be specific about what to create/modify/audit."
+        },
+        "type": {
+          "type": "string",
+          "enum": [
+            "create",
+            "audit",
+            "token"
+          ],
+          "description": "Agent type. Defaults to \"create\" if omitted."
+        }
+      },
+      "required": [
+        "prompt"
+      ]
+    }
+  }
+]