AIGE

Check Generation Task Status

antix_check_task

Check the status of an in-progress generation task (returned by antix_generate_image or antix_generate_video). Returns one of: queued, processing, completed, failed.

🚨 CRITICAL: SHOW THE RESULT ONCE — NEVER TWICE. When status: completed, the response carries a markdownEmbed string and an instruction field.

• If an AIGE panel/widget rendered this result in the conversation (MCP Apps hosts: Claude, ChatGPT, Cursor) the user can already SEE it — add commentary only. Pasting the markdown as well is a duplicate, and duplicates are a bug. • Otherwise paste markdownEmbed EXACTLY VERBATIM at the top of your reply, exactly once. Example:

Here is your image:

![Generated image 1](https://aige.ws/v/<uid>/<characterId>/<historyId>/raw.png)

<your commentary about the result>

The markdown image syntax ![alt](url) is what makes the image render INLINE in a markdown chat client. Without it — and without a panel — the user sees only prose and no picture.

• The URLs in markdownEmbed are already AIGE links (aige.ws/v/…/raw.png, which redirects to the file). Use them verbatim. Do NOT swap in a raw firebasestorage.googleapis.com URL: those expose the storage path and a download token in the conversation. • DO NOT replace the markdown with a description. Describe AFTER showing, not instead. • DO NOT paste markdownEmbed more than once in a reply. • For multiple images the string already concatenates them all — paste the whole thing once. • For video tasks it renders the poster frame AND a Play link.

Status branches: queued / processing → tell the user 'still working, polling again in 10s' and call this tool again. completed → paste markdownEmbed first, then commentary. failed → tell the user, surface error. Credits were auto-refunded (refundedCredits).

Response shape (completed image task): status, progress, images[].url (direct file), images[].viewUrl (AIGE page), markdownEmbed, instruction (a directive to you), creditsSpent, model, characterId, historyId. Response shape (completed video task): status, progress, videoUrl (direct file), viewUrl (AIGE page), posterUrl, markdownEmbed (poster + play link), instruction, creditsSpent, model, characterId, historyId.

Polling cadence: Images typically complete in 10-25s. Videos take 30-90s. Wait ~10s then call once for images; for videos, ~30s then every 15s. (On MCP Apps hosts an AIGE panel renders with this tool and keeps polling by itself — videos play inline there with a real player — so let the panel show it and comment only; paste markdownEmbed on clients WITHOUT the panel.)

Input Schema

Name	Required	Description	Default
`taskId`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

First observedOct 21, 2026

TDQS

A4.5/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden and does so thoroughly: it discloses refund behavior on failure, credit spending, latency expectations (10-25s images, 30-90s videos), and the full response shape for both image and video tasks. It also flags the URL-safety constraint (never substitute raw firebasestorage URLs) and the panel-vs-no-panel rendering distinction, which are behavioral traits no schema could express.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Purpose is front-loaded and the status branches are well-ordered, but the response is heavily padded: the 'never show twice' rule is restated four separate times (pasting once, not replacing with a description, not pasting twice, whole concatenated string), and emoji/caps formatting inflates length. Much of the presentation directive is genuinely necessary, but the repetition costs it.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

There is no output schema, so the description must document return values itself — and it does, with separate response shapes for completed image and video tasks listing every field. Combined with branching behavior, refunds, and polling timing, an agent has everything needed to call and handle the result correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0% for the single taskId parameter, but the description compensates by explaining where the value comes from ('returned by antix_generate_image or antix_generate_video'). It gives origin and semantics but not format, so slightly below the no-parameter baseline of 4-plus.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Check the status of an in-progress generation task') and immediately disambiguates by naming the producing tools (antix_generate_image / antix_generate_video). It even enumerates the possible return values (queued, processing, completed, failed), so an agent knows exactly what question this tool answers.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives explicit prescriptive guidance per status branch ('queued/processing → poll again in 10s', 'completed → paste markdownEmbed first', 'failed → surface error, credits refunded') plus concrete polling cadence for images vs videos. The one gap is that the adjacent sibling antix_check_batch is never mentioned, so an agent holding a batch task has no stated rule for which status tool to call.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All AIGE tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?