← EmpirioLabs AICONTENT HISTORY

Update to EmpirioLabs AI

Snapshot Nov 16, 2026 · 00:09 UTC · version 1.0.13

Collection source: downloaded plugin package.

WHAT CHANGED

Instructions updated for get-started

Skill instructions — Now says: 5. Put every output link in the reply so the user can open the file. Show the file through its link only. Write no image or video element in the reply, not...

Observed in instructions or declared skills. Runtime behavior has not been tested.

Save EmpirioLabs AI and check future changes in Following.

Up to 3 plugins free · no card required.

All fields and original values.

Skill instructions

Before

Nov 15, 2026 · 06:36 UTC

5. Put every output link in the reply so the user can open the file.

After

Nov 16, 2026 · 00:09 UTC

5. Put every output link in the reply so the user can open the file. Show the file through its link only. Write no image or video element in the reply, not even an empty one (no AsyncImage, no AsyncVideo): such an element shows media fou...

Compare saved observations

Download comparison JSON

SKILL.md text changes

− removed · + added

--- before
+++ after
@@ -10,6 +10,6 @@
 2. Offer a first thing to try and wait for the choice. Good first requests: an image from a one-line description, a short text read aloud, or a five-second video.
 3. Run the request with the matching tool: `generate_image`, `text_to_speech` (choose a voice with `list_voices`), or `generate_video`.
 4. When a result is still generating, call `get_job` with its job id, and again until the output is ready. Never start the same generation twice.
-5. Put every output link in the reply so the user can open the file.
+5. Put every output link in the reply so the user can open the file. Show the file through its link only. Write no image or video element in the reply, not even an empty one (no AsyncImage, no AsyncVideo): such an element shows media found by a search or an empty frame, never the generated file.
 
 If the user asks which model to use, call `list_models` for that kind of output and suggest one. If a request is refused, repeat the reason the tool gave and say what to change.
Full technical diff · 1 changed fields

changed /skill_md_contents

BEFORE
"---\nname: get-started\ndescription: Get started with EmpirioLabs AI in ChatGPT. Use right after the app is connected, or when the user asks what the app can do or how to begin.\n---\n# Get started with EmpirioLabs AI\n\nEmpirioLabs AI runs image, video, speech, music, 3D, transcription, document and text models on the user's own EmpirioLabs account.\n\n1. Say in two or three lines what the app can make: images and image edits, short videos, spoken audio in many voices, music and sound effects, 3D models from a picture, transcripts of recordings, Markdown from PDFs and scans, web search with sources, and answers from text models.\n2. Offer a first thing to try and wait for the choice. Good first requests: an image from a one-line description, a short text read aloud, or a five-second video.\n3. Run the request with the matching tool: `generate_image`, `text_to_speech` (choose a voice with `list_voices`), or `generate_video`.\n4. When a result is still generating, call `get_job` with its job id, and again until the output is ready. Never start the same generation twice.\n5. Put every output link in the reply so the user can open the file.\n\nIf the user asks which model to use, call `list_models` for that kind of output and suggest one. If a request is refused, repeat the reason the tool gave and say what to change.\n"
AFTER
"---\nname: get-started\ndescription: Get started with EmpirioLabs AI in ChatGPT. Use right after the app is connected, or when the user asks what the app can do or how to begin.\n---\n# Get started with EmpirioLabs AI\n\nEmpirioLabs AI runs image, video, speech, music, 3D, transcription, document and text models on the user's own EmpirioLabs account.\n\n1. Say in two or three lines what the app can make: images and image edits, short videos, spoken audio in many voices, music and sound effects, 3D models from a picture, transcripts of recordings, Markdown from PDFs and scans, web search with sources, and answers from text models.\n2. Offer a first thing to try and wait for the choice. Good first requests: an image from a one-line description, a short text read aloud, or a five-second video.\n3. Run the request with the matching tool: `generate_image`, `text_to_speech` (choose a voice with `list_voices`), or `generate_video`.\n4. When a result is still generating, call `get_job` with its job id, and again until the output is ready. Never start the same generation twice.\n5. Put every output link in the reply so the user can open the file. Show the file through its link only. Write no image or video element in the reply, not even an empty one (no AsyncImage, no AsyncVideo): such an element shows media found by a search or an empty frame, never the generated file.\n\nIf the user asks which model to use, call `list_models` for that kind of output and suggest one. If a request is refused, repeat the reason the tool gave and say what to change.\n"
Full snapshot data
{
  "description": "Get started with EmpirioLabs AI in ChatGPT. Use right after the app is connected, or when the user asks what the app can do or how to begin.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 202
    }
  ],
  "name": "get-started",
  "skill_md_contents": "---\nname: get-started\ndescription: Get started with EmpirioLabs AI in ChatGPT. Use right after the app is connected, or when the user asks what the app can do or how to begin.\n---\n# Get started with EmpirioLabs AI\n\nEmpirioLabs AI runs image, video, speech, music, 3D, transcription, document and text models on the user's own EmpirioLabs account.\n\n1. Say in two or three lines what the app can make: images and image edits, short videos, spoken audio in many voices, music and sound effects, 3D models from a picture, transcripts of recordings, Markdown from PDFs and scans, web search with sources, and answers from text models.\n2. Offer a first thing to try and wait for the choice. Good first requests: an image from a one-line description, a short text read aloud, or a five-second video.\n3. Run the request with the matching tool: `generate_image`, `text_to_speech` (choose a voice with `list_voices`), or `generate_video`.\n4. When a result is still generating, call `get_job` with its job id, and again until the output is ready. Never start the same generation twice.\n5. Put every output link in the reply so the user can open the file. Show the file through its link only. Write no image or video element in the reply, not even an empty one (no AsyncImage, no AsyncVideo): such an element shows media found by a search or an empty frame, never the generated file.\n\nIf the user asks which model to use, call `list_models` for that kind of output and suggest one. If a request is refused, repeat the reason the tool gave and say what to change.\n"
}

SHA-256 of public snapshot: 28b5093b73d99225897772e4aac7cdd07a8f78b38901610e84cb1d365a389460