Bonsai-27B-GGUF
lmstudio-community · Kit
A GGUF conversion of the 27B Bonsai model for LM Studio, with a 3.80 GB Q1_0 language file and a 0.93 GB BF16 vision projector. Tested for text, vision and par…
Shows a traditional RPA script and the ANCHOR automation agent running on the same web page before and after it changes, at five levels of change, in your browser.
ANCHOR: the page changed, which robot noticed? is a kit in the Developer category, listed by Mindket. It is free. This is version 1, last updated on . Download it on this page, a zip of its files. Nothing in it runs by itself.
pip install -U huggingface_hub hf download isntlucxs/anchor-demo --repo-type space --local-dir anchor-demo cd anchor-demo
This Space sets two kinds of automation against one web page that gets altered. The first is an old-style RPA script that repeats element selectors saved from the unaltered page. The second is ANCHOR, an open-source agent for web tasks that works on your own machine, and here it runs an automation it saved earlier and uses no AI.
People who build or buy browser automation and want to see how it copes when a page changes, and developers curious about how ANCHOR confirms its own steps.
| Industry | Who uses it | How they use it | Example |
|---|---|---|---|
| Software testing | QA engineers | Choose a task and a change level, then compare the two robots | Seeing which robot survives renamed ids |
| Process automation | RPA developers | Watch a selector script run against the altered page | Checking where a recorded script breaks |
| Security | Security reviewers | Try the top change level, which slips malicious content into the page | Seeing what each robot does with it |
Skip it if you want to build automations here, since this Space only replays recorded runs.
The Space runs in these places.
| Use case | Who | What they get back |
|---|---|---|
| Side-by-side comparison | Anyone curious | The live page in its original and altered states, plus a record of each robot's actions |
| Pick a difficulty | Testers | A scale of five levels, starting with changed element names and ending with malicious content slipped into the page |
| Read the benchmark results | Researchers | Measured outcomes for the full benchmark, listed with the models used |
Browsing needs only a web browser. A local copy needs a few more things first.
pip install -U huggingface_hub hf download isntlucxs/anchor-demo --repo-type space --local-dir anchor-demo cd anchor-demo
Choose a task, then choose how far the page should differ from the original. The page lays out both states and what each robot did in them. With a local copy, go into the anchor-demo folder and open index.html in your browser.
cd anchor-demo
| Tool | What it does |
|---|---|
| Task picker | Chooses which task's page you look at |
| Change level | Sets how far the page differs from the original, on a scale of five |
| RPA script | Repeats the saved element selectors from the unchanged page |
| ANCHOR | Runs a previously saved automation without any AI |
You pick a task and set the change to the level with renamed ids. The page shows it before and after the change, plus what the script and ANCHOR each did on it.
If you are new, RPA means a script that repeats clicks it recorded, and it can break when a page changes. Pick one task, start at the lowest change level and compare the two robots.
If you know the field, the test pages belong to ANCHOR's resilience benchmark. The runs were captured in advance by version 0.5.0 together with the benchmark's code. Because no AI is involved, repeating a task always gives an identical outcome. In ANCHOR a small local model makes the plan, while ordinary code does the clicking and confirms every step took effect.
The Space shows recorded runs, so it does not run ANCHOR for you.
Shows a traditional RPA script and the ANCHOR automation agent running on the same web page before and after it changes, at five levels of change, in your browser.
Made by isntlucxs on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.
What it is: This Space sets two kinds of automation against one web page that gets altered. The first is an old-style RPA script that repeats element selectors saved from the unaltered page. The second is ANCHOR, an open-source agent for web tasks that works on your own machine, and here it runs an automation it saved earlier and uses no AI.
Who it suits: People who build or buy browser automation and want to see how it copes when a page changes, and developers curious about how ANCHOR confirms its own steps.
Industry | Who uses it | How they use it | Example
Software testing | QA engineers | Choose a task and a change level, then compare the two robots | Seeing which robot survives renamed ids
Process automation | RPA developers | Watch a selector script run against the altered page | Checking where a recorded script breaks
Security | Security reviewers | Try the top change level, which slips malicious content into the page | Seeing what each robot does with it
Skip it if you want to build automations here, since this Space only replays recorded runs.
Where to use it: The Space runs in these places.
Use case | Who | What they get back
Side-by-side comparison | Anyone curious | The live page in its original and altered states, plus a record of each robot's actions
Pick a difficulty | Testers | A scale of five levels, starting with changed element names and ending with malicious content slipped into the page
Read the benchmark results | Researchers | Measured outcomes for the full benchmark, listed with the models used
How to set it up: Browsing needs only a web browser. A local copy needs a few more things first.
pip install -U huggingface_hub
hf download isntlucxs/anchor-demo --repo-type space --local-dir anchor-demo
cd anchor-demo
How to use it: Choose a task, then choose how far the page should differ from the original. The page lays out both states and what each robot did in them. With a local copy, go into the anchor-demo folder and open index.html in your browser.
cd anchor-demo
Tool | What it does
Task picker | Chooses which task's page you look at
Change level | Sets how far the page differs from the original, on a scale of five
RPA script | Repeats the saved element selectors from the unchanged page
ANCHOR | Runs a previously saved automation without any AI
Example: You pick a task and set the change to the level with renamed ids. The page shows it before and after the change, plus what the script and ANCHOR each did on it.
For newcomers and experts: If you are new, RPA means a script that repeats clicks it recorded, and it can break when a page changes. Pick one task, start at the lowest change level and compare the two robots.
If you know the field, the test pages belong to ANCHOR's resilience benchmark. The runs were captured in advance by version 0.5.0 together with the benchmark's code. Because no AI is involved, repeating a task always gives an identical outcome. In ANCHOR a small local model makes the plan, while ordinary code does the clicking and confirms every step took effect.
Pick a task and a change level
The page loads before and after the change
The RPA script repeats its selectors
ANCHOR runs its saved automation
You compare what each one did
Good to know: The Space shows recorded runs, so it does not run ANCHOR for you.
Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.
README.md 5 KB
Listed Hugging Face Spaces in the same category that share the most tags with this one.
| Product | Likes | Built with | License | Last change |
|---|---|---|---|---|
| ANCHOR: the page changed, which robot noticed? This page | 1 | static | apache-2.0 | |
| InferenceSupportShares static | 764 | static | afl-3.0 | |
| HuggyShares static | 463 | static | cc-by-nc-sa-4.0 | |
| Z-Image Lookalike LoRA IndexShares static | 189 | static | apache-2.0 | |
| Granite Docling 258M WebGPUShares static | 163 | static | apache-2.0 | |
| DINOv3 WebShares static | 162 | static | apache-2.0 |
lmstudio-community · Kit
A GGUF conversion of the 27B Bonsai model for LM Studio, with a 3.80 GB Q1_0 language file and a 0.93 GB BF16 vision projector. Tested for text, vision and par…
LibertAIDAI · Kit
Runs DeepSeek-V4.1-Flash-1017 as a 2.9 bits-per-weight EXL3 quantization, a 196 GiB copy for people serving large models on their own hardware. Perplexity stay…
facebook · Kit
Generates short music clips from a text description or an audio prompt, in the small checkpoint of facebook's MusicGen. It suits developers and music hobbyists…
alpindale · Kit
Tags every character of C source code for syntax highlighting with a tiny neural network of about 984 weights, run through a small C program. For developers an…
prithivMLmods · Kit
Try a set of LoRAs for the FLUX.2 Klein image model in a Gradio app that runs in your browser or on your own computer. Made for people who make or test AI imag…
Sokheng · Kit
Reads Khmer and English text from documents, receipts and photos of signs, using a detector for text lines and a recogniser for each line. Suits developers who…