Open-source platform for 3D AI agents. Turn text or a photo into a rigged, animated GLB avatar, give it an LLM brain, memory and a wallet, and embed it anywhere with one web component. Remote MCP server, x402 payments, Three.js, browser-native.
JavaScript 3d-generationai-agentsanimationavatardigital-humangenerative-aiglbgltfimage-to-3dllmmcpmcp-servermodel-context-protocolsolanatext-to-3dthreejsweb-componentswebglwebxrx402
fix(mcp-studio): check_job writes, so stop declaring it a read-only status probe
nirholas042eada52e5f
diff --git a/CHANGELOG.md b/CHANGELOG.mdindex 545fed69a..c460c9480 100644--- a/CHANGELOG.md+++ b/CHANGELOG.md@@ -4,6 +4,10 @@Public history for [three.ws](https://three.ws), newest first. New pages come from `added` dates in data/pages.json; everything else is curated in data/changelog.json. Also available as [JSON](https://three.ws/changelog.json) and [RSS](https://three.ws/changelog.xml), live at [three.ws/changelog](https://three.ws/changelog).+## 2026-09-12++- **Corrected what the 3D Studio connector declares about checking a pending job**: The tool that collects a finished 3D generation was labelled a read-only status probe. It is not: the first check that finds a job done copies the model into permanent storage, records it, and scores it, and it can route failed work to another provider. Any AI client reading those labels to decide whether a call is safe to repeat was being told the wrong thing. The labels and the documentation now match the behaviour. Nothing about how the tool works changed. (`/docs/mcp-studio`) `[fix, sdk]`+## 2026-09-11- **Selfie to avatar is unstuck: a scan that used to wait forever now finishes in seconds**: Uploading a photo on the selfie page could sit on the same message indefinitely and never start the scan, with no error and nothing to retry. One copy of the face-tracking runtime held at our CDN was telling browsers to expect 11 MB and then sending a third of that, so the page waited for bytes that were never coming. That copy is gone, and three changes make the failure impossible to repeat silently: the face-tracking runtime now loads from three.ws first instead of a public CDN, any wait on it gives up after fifteen seconds and builds your avatar from the photo as it is, and a photo is never rejected for having no face in it just because the detector could not load. A new check also walks every large file on the site the way a browser asks for it and fails loudly when the bytes do not add up. Measured after the fix: a selfie became a rigged, animation-ready avatar in about fifteen seconds, and still finished in under a minute with the face-tracking runtime completely unreachable. (`/create/selfie`) `[fix, infra]`diff --git a/api/_mcp-studio/tools.js b/api/_mcp-studio/tools.jsindex 1f7e42ef6..848c46d5f 100644--- a/api/_mcp-studio/tools.js+++ b/api/_mcp-studio/tools.js@@ -980,7 +980,8 @@ const DEFS = [description:'Check on a 3D generation that returned status "pending" and collect the finished model. Pass the ' +'job_id from the pending result. While it is still rendering you get updated timing; call again after ' +- 'the suggested wait. When it is done the model renders inline in the interactive 3D viewer.',+ 'the suggested wait. Checking can save the completed model and restart failed work on another ' ++ 'provider. When it is done the model renders inline in the interactive 3D viewer.',inputSchema: {type: 'object',additionalProperties: false,@@ -995,9 +996,9 @@ const DEFS = [},},annotations: {- readOnlyHint: true, // a status probe; creates nothing+ readOnlyHint: false, // polling can persist a model and submit recovery jobsdestructiveHint: false,- idempotentHint: true, // same job_id, same answer until the job advances+ idempotentHint: false, // later checks can advance recovery to another provideropenWorldHint: true,},_meta: widgetMeta('Checking your 3D model…', 'Here is your 3D model'),diff --git a/data/changelog.json b/data/changelog.jsonindex d997cbff3..d3910f9cf 100644--- a/data/changelog.json+++ b/data/changelog.json@@ -28146,6 +28146,16 @@"sdk"],"link": "/docs/agent-skills"+ },+ {+ "date": "2026-09-12",+ "title": "Corrected what the 3D Studio connector declares about checking a pending job",+ "summary": "The tool that collects a finished 3D generation was labelled a read-only status probe. It is not: the first check that finds a job done copies the model into permanent storage, records it, and scores it, and it can route failed work to another provider. Any AI client reading those labels to decide whether a call is safe to repeat was being told the wrong thing. The labels and the documentation now match the behaviour. Nothing about how the tool works changed.",+ "tags": [+ "fix",+ "sdk"+ ],+ "link": "/docs/mcp-studio"}]}diff --git a/docs/mcp-studio.md b/docs/mcp-studio.mdindex 6103d3ac8..cf17bcf23 100644--- a/docs/mcp-studio.md+++ b/docs/mcp-studio.md@@ -71,9 +71,14 @@ All eleven are free and keyless.The six generation tools run operator-funded on the platform's own generationpipeline. Annotations: `readOnlyHint:false`, `destructiveHint:false`,`idempotentHint:false`, `openWorldHint:true` (work runs against external model-APIs; nothing is ever modified or deleted). `check_job` is the exception: it is a-pure status probe (`readOnlyHint:true`, `idempotentHint:true`) and never counts-against the generation quota. `look_at_model` is read-only and idempotent too,+APIs; nothing is ever modified or deleted). `check_job` reads like a status probe+but is not one, and its annotations say so (`readOnlyHint:false`,+`idempotentHint:false`): the FIRST check that finds a job finished materializes+the creation, copying the model into our storage, recording the row and running+the quality gate, and it can route failed work to another provider. Later checks+of the same job are served from the done-frame cache, which is exactly why it is+not idempotent either. It still never counts against the generation quota.+`look_at_model` is genuinely read-only and idempotent,but it renders frames server-side, so it rides the same per-IP generationquota as the six generators.diff --git a/prompts/store-submissions/_generated/chatgpt-app-submission.json b/prompts/store-submissions/_generated/chatgpt-app-submission.jsondeleted file mode 100644index c54963961..000000000--- a/prompts/store-submissions/_generated/chatgpt-app-submission.json+++ /dev/null@@ -1,212 +0,0 @@-{- "$schema": "https://developers.openai.com/apps-sdk/schemas/chatgpt-app-submission.v1.json",- "schema_version": 1,- "app_info": {- "display_name": "three.ws 3D Studio",- "subtitle": "Create 3D models from text",- "description": "Describe any object or character and three.ws 3D Studio builds a real, textured 3D model, then shows it in an interactive viewer inside the conversation. Eleven tools cover the path from idea to asset: generate a model from text or a reference image, generate an avatar, auto-rig a static model so it can be animated, refine a model by describing a change, and inspect a finished model from several angles. Every result downloads as a standard GLB that opens in Blender, Unity, Unreal, three.js, or any glTF pipeline. No account, no API key, no payment.",- "category": "DESIGN"- },- "tools": {- "forge_free": {- "annotations": {- "readOnlyHint": false,- "openWorldHint": true,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Runs a 3D generation and writes a new GLB file into our object storage, then returns that file's URL.",- "open_world_justification": "Publishes the generated GLB at a public URL anyone with the link can fetch, and relies on third-party inference providers to produce it.",- "destructive_justification": "Only adds a new file and never overwrites or deletes an existing one, so a model supplied as input is left unchanged."- }- },- "text_to_avatar": {- "annotations": {- "readOnlyHint": false,- "openWorldHint": true,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Runs a 3D generation and writes a new GLB file into our object storage, then returns that file's URL.",- "open_world_justification": "Publishes the generated GLB at a public URL anyone with the link can fetch, and relies on third-party inference providers to produce it.",- "destructive_justification": "Only adds a new file and never overwrites or deletes an existing one, so a model supplied as input is left unchanged."- }- },- "mesh_forge": {- "annotations": {- "readOnlyHint": false,- "openWorldHint": true,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Runs a 3D generation and writes a new GLB file into our object storage, then returns that file's URL.",- "open_world_justification": "Publishes the generated GLB at a public URL anyone with the link can fetch, and relies on third-party inference providers to produce it.",- "destructive_justification": "Only adds a new file and never overwrites or deletes an existing one, so a model supplied as input is left unchanged."- }- },- "rig_mesh": {- "annotations": {- "readOnlyHint": false,- "openWorldHint": true,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Runs a 3D generation and writes a new GLB file into our object storage, then returns that file's URL.",- "open_world_justification": "Publishes the generated GLB at a public URL anyone with the link can fetch, and relies on third-party inference providers to produce it.",- "destructive_justification": "Only adds a new file and never overwrites or deletes an existing one, so a model supplied as input is left unchanged."- }- },- "forge_avatar": {- "annotations": {- "readOnlyHint": false,- "openWorldHint": true,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Runs a 3D generation and writes a new GLB file into our object storage, then returns that file's URL.",- "open_world_justification": "Publishes the generated GLB at a public URL anyone with the link can fetch, and relies on third-party inference providers to produce it.",- "destructive_justification": "Only adds a new file and never overwrites or deletes an existing one, so a model supplied as input is left unchanged."- }- },- "refine_model": {- "annotations": {- "readOnlyHint": false,- "openWorldHint": true,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Runs a 3D generation and writes a new GLB file into our object storage, then returns that file's URL.",- "open_world_justification": "Publishes the generated GLB at a public URL anyone with the link can fetch, and relies on third-party inference providers to produce it.",- "destructive_justification": "Only adds a new file and never overwrites or deletes an existing one, so a model supplied as input is left unchanged."- }- },- "check_job": {- "annotations": {- "readOnlyHint": true,- "openWorldHint": true,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Looks up an existing generation job by its id and reports that job's state without creating or changing anything.",- "open_world_justification": "Reports on work running on third-party inference providers, so its result reflects external systems rather than data we control.",- "destructive_justification": "Only reads job state and never deletes, overwrites or cancels anything."- }- },- "look_at_model": {- "annotations": {- "readOnlyHint": true,- "openWorldHint": true,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Renders a model that already exists from several angles and returns the frames as images, storing nothing and changing nothing.",- "open_world_justification": "Fetches an arbitrary public GLB URL, so it reaches third-party hosts outside our own domain.",- "destructive_justification": "Only reads the supplied model in order to render it, and never modifies or deletes the source file."- }- },- "create_agent_persona": {- "annotations": {- "readOnlyHint": false,- "openWorldHint": true,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Saves a new persona record and copies the model into our durable storage so the body outlives the source URL.",- "open_world_justification": "Stores the persona privately in our own database but publishes its model at a public URL, and calls external model and speech providers.",- "destructive_justification": "Only creates a new persona and never modifies or deletes an existing persona or model."- }- },- "get_agent_persona": {- "annotations": {- "readOnlyHint": true,- "openWorldHint": false,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Looks up a stored persona by its id and returns that persona's configuration without creating or changing anything.",- "open_world_justification": "Reads only persona records held in our own private store and contacts no external system.",- "destructive_justification": "Only reads a persona record and never writes or removes anything."- }- },- "persona_say": {- "annotations": {- "readOnlyHint": false,- "openWorldHint": false,- "destructiveHint": false- },- "justifications": {- "read_only_justification": "Increments the persona's turn counter in our own store while returning the render directive for this turn.",- "open_world_justification": "Acts only on a persona in our own private store and renders through our own embed, reaching no external system.",- "destructive_justification": "Only appends a turn and updates a counter, never deleting or overwriting the persona's configuration or model."- }- }- },- "test_cases": [- {- "description": "Generate a 3D prop from a text description",- "user_prompt": "Make a 3D model of a friendly round robot mascot, glossy white plastic.",- "file_attachment_urls": null,- "tools_triggered": "forge_free, then check_job if the first response returns status \"pending\"",- "expected_output": "An inline interactive 3D viewer with the textured model, plus Download, Spin, Recenter and Open in three.ws. It auto-rotates until dragged. Free, no account. Takes one to four minutes; past the inline wait it returns a job_id and check_job collects the finished model.",- "expected_output_url": null- },- {- "description": "Generate a rigged, animation-ready character",- "user_prompt": "Make a rigged, animation-ready knight character I can pose.",- "file_attachment_urls": null,- "tools_triggered": "forge_avatar, then check_job if the first response returns status \"pending\"",- "expected_output": "A rigged GLB in the same inline viewer, with a humanoid skeleton and skin weights already applied, so an idle animation plays rather than the model standing in a bind pose. One call performs both the mesh generation and the rig.",- "expected_output_url": null- },- {- "description": "Iterate on a model by describing the change in words",- "user_prompt": "Now make that robot's shell matte instead of glossy.",- "file_attachment_urls": null,- "tools_triggered": "refine_model",- "expected_output": "Run immediately after test case 1, in the same conversation. Returns a new version anchored to the previous model rather than an unrelated regeneration, and replaces the GLB in the viewer. The earlier version stays reachable, so the change can be reverted.",- "expected_output_url": null- },- {- "description": "Let the assistant see a 3D model and report on it",- "user_prompt": "Look at this 3D model and tell me what it is and whether it has any obvious defects: https://three.ws/cdn/objects/polyhaven/glb/ArmChair_01.glb",- "file_attachment_urls": null,- "tools_triggered": "look_at_model",- "expected_output": "Frames from several angles returned as MCP image content blocks, which the client renders into the conversation, plus geometry stats. The assistant describes the armchair from images it can actually see. Free, seconds not minutes. The URL is ours and needs no credentials.",- "expected_output_url": null- },- {- "description": "Turn a generated model into a persistent agent body that speaks",- "user_prompt": "Save that knight as a persistent agent body called Sir Gareth, then have him introduce himself.",- "file_attachment_urls": null,- "tools_triggered": "create_agent_persona, then persona_say",- "expected_output": "Run after test case 2, in the same conversation. An inline living-body widget framed from https://three.ws, in which the avatar lip-syncs the line with a matching expression and gesture. Returns a persona_id that get_agent_persona reloads in a fresh session.",- "expected_output_url": null- }- ],- "negative_test_cases": [- {- "description": "The user wants a 2D image, not a 3D model. The wording nearly matches our main generation prompt, but this plugin only returns 3D GLB files, so image generation should handle it.",- "user_prompt": "Draw a cartoon robot mascot for my startup's landing page.",- "file_attachment_urls": null,- "tools_triggered": null,- "expected_output": "The app should not be invoked, because it only produces 3D GLB models and the user asked for a 2D image.",- "expected_output_url": null- },- {- "description": "\"Model\" is a verb here, meaning a financial projection. Nothing 3D is involved, and no tool in this plugin operates on spreadsheets or forecasts.",- "user_prompt": "Model out our Q3 revenue if we raise prices 20%.",- "file_attachment_urls": null,- "tools_triggered": null,- "expected_output": "The app should not be invoked, because no tool here operates on spreadsheets, forecasts or financial data.",- "expected_output_url": null- },- {- "description": "The user asks how to do something in other software, not for work on a file. It says \"rig\" and \"character\", but rig_mesh needs the URL of an existing GLB and would return nothing they asked for.",- "user_prompt": "How do I rig a humanoid character in Blender?",- "file_attachment_urls": null,- "tools_triggered": null,- "expected_output": "The app should not be invoked, because the user wants an explanation of other software rather than work on a file.",- "expected_output_url": null- }- ]-}diff --git a/prompts/store-submissions/_generated/openai-portal-fields.md b/prompts/store-submissions/_generated/openai-portal-fields.mdindex 22497260d..8a3d1d171 100644--- a/prompts/store-submissions/_generated/openai-portal-fields.md+++ b/prompts/store-submissions/_generated/openai-portal-fields.md@@ -108,19 +108,19 @@ Only adds a new file and never overwrites or deletes an existing one, so a model### check_job-**Read Only** (113/200)+**Read Only** (144/200)```-Looks up an existing generation job by its id and reports that job's state without creating or changing anything.+The first check that finds a job finished copies the model into our storage and records the creation, so the call writes rather than only reads.```-**Open World** (128/200)+**Open World** (129/200)```-Reports on work running on third-party inference providers, so its result reflects external systems rather than data we control.+Collects work from third-party inference providers and publishes the finished GLB at a public URL anyone with the link can fetch.```-**Destructive** ( 71/200)+**Destructive** ( 76/200)```-Only reads job state and never deletes, overwrites or cancels anything.+Only adds the finished model and never deletes or overwrites an earlier one.```### look_at_model@@ -149,7 +149,7 @@ Saves a new persona record and copies the model into our durable storage so the**Open World** (136/200)```-Stores the persona privately in our own database but publishes its model at a public URL, and calls external model and speech providers.+Fetches the GLB from a public URL outside our domain and republishes it from our storage at a public URL anyone with the link can fetch.```**Destructive** ( 86/200)@@ -235,12 +235,12 @@ Make a rigged, animation-ready knight character I can pose.**Tool triggered**```-forge_avatar, then check_job if the first response returns status "pending"+forge_avatar, then check_job if it returns status "pending", then rig_mesh if the finished job was only the mesh```-**Expected output** (228/300)+**Expected output** (257/300)```-A rigged GLB in the same inline viewer, with a humanoid skeleton and skin weights already applied, so an idle animation plays rather than the model standing in a bind pose. One call performs both the mesh generation and the rig.+A GLB in the inline viewer with a humanoid skeleton and skin weights applied, so it can be posed. One call normally does mesh and rig; if it times out at the mesh stage the pending result says to finish with rig_mesh. A humanoid rig also plays an idle clip.```### Test Case 3@@ -257,7 +257,7 @@ Now make that robot's shell matte instead of glossy.**Tool triggered**```-refine_model+refine_model, then check_job if it returns status "pending"```**Expected output** (256/300)diff --git a/public/changelog-recent.json b/public/changelog-recent.jsonindex fb7e531db..d8bda51e0 100644--- a/public/changelog-recent.json+++ b/public/changelog-recent.json@@ -1,14 +1,25 @@{- "generated_at": "2026-09-11T23:57:28.226Z",+ "generated_at": "2026-09-12T01:51:12.338Z","generated_by": "scripts/build-page-index.mjs from data/pages.json + data/changelog.json","site": {"name": "three.ws","url": "https://three.ws"},- "total": 3673,+ "total": 3674,"complete": false,"full_feed": "/changelog.json","entries": [+ {+ "date": "2026-09-12",+ "type": "update",+ "title": "Corrected what the 3D Studio connector declares about checking a pending job",+ "summary": "The tool that collects a finished 3D generation was labelled a read-only status probe. It is not: the first check that finds a job done copies the model into permanent storage, records it, and scores it, and it can route failed work to another provider. Any AI client reading those labels to decide whether a call is safe to repeat was being told the wrong thing. The labels and the documentation now match the behaviour. Nothing about how the tool works changed.",+ "link": "/docs/mcp-studio",+ "tags": [+ "fix",+ "sdk"+ ]+ },{"date": "2026-09-11","type": "update",@@ -1609,16 +1620,6 @@"tags": ["fix"]- },- {- "date": "2026-09-07",- "type": "update",- "title": "Robinhood Portfolios: describe a portfolio in a sentence, hold it on-chain",- "summary": "Robinhood Chain is the one place where tokenized stocks, crypto majors and memecoins trade against each other in the same pools, so a basket holding NVDA, WETH and a launchpad coin is just three tokens in one vault. Write a sentence at three.ws/markets/robinhood/portfolios and a screen ranks all 700-odd holdable assets against it, then hands back a real basket with target weights, a thesis, and a per-holding reason for every pick. The whole methodology, the prompt, the weights, the schedule, and every holding's price and liquidity at the moment it was chosen, is hashed into a manifest you can download, so the document you read is provably the one the portfolio is deployed against. The vault behind it holds the basket in kind: you deliver every holding to mint and receive every holding to redeem, so it never sells anything to let somebody in or out and needs no price feed to work at all, which matters because most of this chain has none. Rebalancing back to target happens on the schedule the manifest published, and is offered as a descending-price auction anyone can fill rather than a trade a manager makes. Refining a portfolio mints a new version pointing back at its parent instead of editing what you already hold, and forks pay their ancestors out of a fixed slice of fees. When a tokenized stock is paused by its issuer everything stops until it resumes, and the page says so, because a halted share cannot be moved at any price. A companion board at /markets/robinhood/portfolios/universe lists every asset a portfolio may hold, with live liquidity, volume, and the premium each stock trades at against its Chainlink price.",- "link": "/markets/robinhood/portfolios",- "tags": [- "feature"- ]}]}diff --git a/public/changelog.json b/public/changelog.jsonindex 1f95140f1..fb7cc60c6 100644--- a/public/changelog.json+++ b/public/changelog.json@@ -1,11 +1,22 @@{- "generated_at": "2026-09-11T23:57:28.205Z",+ "generated_at": "2026-09-12T01:51:12.316Z","generated_by": "scripts/build-page-index.mjs from data/pages.json + data/changelog.json","site": {"name": "three.ws","url": "https://three.ws"},"entries": [+ {+ "date": "2026-09-12",+ "type": "update",+ "title": "Corrected what the 3D Studio connector declares about checking a pending job",+ "summary": "The tool that collects a finished 3D generation was labelled a read-only status probe. It is not: the first check that finds a job done copies the model into permanent storage, records it, and scores it, and it can route failed work to another provider. Any AI client reading those labels to decide whether a call is safe to repeat was being told the wrong thing. The labels and the documentation now match the behaviour. Nothing about how the tool works changed.",+ "link": "/docs/mcp-studio",+ "tags": [+ "fix",+ "sdk"+ ]+ },{"date": "2026-09-11","type": "update",diff --git a/public/changelog.xml b/public/changelog.xmlindex 39559a1f8..f114c9ca8 100644--- a/public/changelog.xml+++ b/public/changelog.xml@@ -6,6 +6,14 @@<atom:link href="https://three.ws/changelog.xml" rel="self" type="application/rss+xml"/><description>What's new on three.ws: features, improvements, and releases, newest first.</description><language>en</language>+ <item>+ <title>Corrected what the 3D Studio connector declares about checking a pending job</title>+ <link>https://three.ws/docs/mcp-studio</link>+ <guid isPermaLink="false">2026-09-12:Corrected what the 3D Studio connector declares about checking a pending job</guid>+ <pubDate>Sat, 12 Sep 2026 12:00:00 GMT</pubDate>+ <category>fix</category>+ <description>The tool that collects a finished 3D generation was labelled a read-only status probe. It is not: the first check that finds a job done copies the model into permanent storage, records it, and scores it, and it can route failed work to another provider. Any AI client reading those labels to decide whether a call is safe to repeat was being told the wrong thing. The labels and the documentation now match the behaviour. Nothing about how the tool works changed.</description>+ </item><item><title>Selfie to avatar is unstuck: a scan that used to wait forever now finishes in seconds</title><link>https://three.ws/create/selfie</link>diff --git a/scripts/build-openai-portal-fields.mjs b/scripts/build-openai-portal-fields.mjsindex 7152f62cb..08a49efbc 100644--- a/scripts/build-openai-portal-fields.mjs+++ b/scripts/build-openai-portal-fields.mjs@@ -35,9 +35,9 @@ const JUSTIFICATIONS = {forge_avatar: GENERATION,refine_model: GENERATION,check_job: {- read_only: 'Looks up an existing generation job by its id and reports that job\'s state without creating or changing anything.',- open_world: 'Reports on work running on third-party inference providers, so its result reflects external systems rather than data we control.',- destructive: 'Only reads job state and never deletes, overwrites or cancels anything.',+ read_only: 'The first check that finds a job finished copies the model into our storage and records the creation, so the call writes rather than only reads.',+ open_world: 'Collects work from third-party inference providers and publishes the finished GLB at a public URL anyone with the link can fetch.',+ destructive: 'Only adds the finished model and never deletes or overwrites an earlier one.',},look_at_model: {read_only: 'Renders a model that already exists from several angles and returns the frames as images, storing nothing and changing nothing.',@@ -46,7 +46,7 @@ const JUSTIFICATIONS = {},create_agent_persona: {read_only: 'Saves a new persona record and copies the model into our durable storage so the body outlives the source URL.',- open_world: 'Stores the persona privately in our own database but publishes its model at a public URL, and calls external model and speech providers.',+ open_world: 'Fetches the GLB from a public URL outside our domain and republishes it from our storage at a public URL anyone with the link can fetch.',destructive: 'Only creates a new persona and never modifies or deletes an existing persona or model.',},get_agent_persona: {@@ -75,14 +75,14 @@ const TEST_CASES = [{scenario: 'Generate a rigged, animation-ready character',prompt: 'Make a rigged, animation-ready knight character I can pose.',- tools: 'forge_avatar, then check_job if the first response returns status "pending"',+ tools: 'forge_avatar, then check_job if it returns status "pending", then rig_mesh if the finished job was only the mesh',expected:- 'A rigged GLB in the same inline viewer, with a humanoid skeleton and skin weights already applied, so an idle animation plays rather than the model standing in a bind pose. One call performs both the mesh generation and the rig.',+ 'A GLB in the inline viewer with a humanoid skeleton and skin weights applied, so it can be posed. One call normally does mesh and rig; if it times out at the mesh stage the pending result says to finish with rig_mesh. A humanoid rig also plays an idle clip.',},{scenario: 'Iterate on a model by describing the change in words',prompt: "Now make that robot's shell matte instead of glossy.",- tools: 'refine_model',+ tools: 'refine_model, then check_job if it returns status "pending"',expected:'Run immediately after test case 1, in the same conversation. Returns a new version anchored to the previous model rather than an unrelated regeneration, and replaces the GLB in the viewer. The earlier version stays reachable, so the change can be reverted.',},@@ -242,6 +242,39 @@ if (process.argv.includes('--json')) {if (!res.ok) throw new Error(`${CONNECTOR} answered ${res.status}`);const served = JSON.parse((await res.text()).trim().split('\n').filter(Boolean).pop().replace(/^data:\s*/, '')).result.tools;+ // The reviewer scans the DEPLOYED server, so the JSON has to describe that.+ // But a local source fix that has not shipped yet would make this file+ // disagree with their scan, which is the same misrepresentation the+ // justifications exist to prevent. Refuse to generate until they match.+ // TOOL_CATALOG / PERSONA_TOOL_CATALOG are the served descriptor arrays; the+ // TOOLS maps beside them are name-keyed handler tables, not descriptors.+ const { TOOL_CATALOG } = await import('../api/_mcp-studio/tools.js');+ const { PERSONA_TOOL_CATALOG } = await import('../api/_mcp-studio/persona-tools.js');+ const local = new Map(+ [...TOOL_CATALOG, ...PERSONA_TOOL_CATALOG].map((t) => [t.name, t.annotations || {}]),+ );+ if (local.size !== served.length) {+ console.error(`this checkout declares ${local.size} tool descriptors but ${CONNECTOR} serves ${served.length}`);+ process.exit(1);+ }+ const drift = [];+ for (const tool of served) {+ const mine = local.get(tool.name);+ if (!mine) continue;+ for (const hint of ['readOnlyHint', 'openWorldHint', 'destructiveHint', 'idempotentHint']) {+ if (mine[hint] !== (tool.annotations || {})[hint]) {+ drift.push(`${tool.name}.${hint}: this checkout says ${mine[hint]}, ${CONNECTOR} serves ${(tool.annotations || {})[hint]}`);+ }+ }+ }+ if (drift.length) {+ console.error('This checkout and the deployed connector disagree about tool annotations:\n');+ for (const d of drift) console.error(` ${d}`);+ console.error('\nThe reviewer scans the deployed server, so deploy this checkout before generating');+ console.error('the submission JSON. Shipping it now would hand them a file that contradicts their own scan.');+ process.exit(1);+ }+const problems = [];const tools = {};for (const tool of served) {diff --git a/tests/fixtures/mcp-golden-tools.json b/tests/fixtures/mcp-golden-tools.jsonindex 646284007..13b279c8a 100644--- a/tests/fixtures/mcp-golden-tools.json+++ b/tests/fixtures/mcp-golden-tools.json@@ -32,9 +32,9 @@"descHead": "dynamic","descHash": "dynamic","annotations": {- "readOnlyHint": true,+ "readOnlyHint": false,"destructiveHint": false,- "idempotentHint": true,+ "idempotentHint": false,"openWorldHint": true},"schemaHash": "06fcfd29d4c4"