Model repo triage
Tom Reilly, ML engineer at BigCo, asked which pull requests were open on the team's model. Memtro listed them with author, comment count and when opened, and read the timeline of the oldest.
Which pull requests are open on our Hugging Face model?
Hugging Face connects Memtro to the Hub. It reads models, datasets and Spaces by author or search with downloads, likes and tags. It also reads a repo's discussions, pull requests and commits, and Jobs. When asked and granted, it can open, comment on, close or rename discussions and pull requests, merge pull requests, create repos, start or cancel paid Jobs, change a Space's hardware, and decide gated access requests. With events on, the Hub tells Memtro when a commit lands or a discussion opens, and a job can start on it. Starting a Job runs code billed per second.
Ask in plain English from chat, your editor, a scheduled job or your own agent. The model picks the right Hugging Face tool, reads what it needs, and answers with the source cited.
These tools change something in Hugging Face. The model is told to use them only when you ask for exactly that, and an administrator grants each one by name to the teams that may use it; a personal connection acts only after its owner allows it.
Turn events on in the connection's Events tab on the Connections page, choose the kinds you want, and Memtro registers its address with Hugging Face for you. Each event can start a job, with the event as its input. Hugging Face sends a secret with every delivery, and one without it is refused.
Events and webhooksOnce Hugging Face is connected, these are the kinds of things people ask. Plain questions, no syntax. If the answer lives partly in another system you have connected, Memtro looks there too.
Which pull requests are open on our model?
Which of our datasets were downloaded most this month?
What changed in the last five commits on main?
Which Jobs failed in our namespace this week?
Comment on discussion 14 that the fix is in the next release.
Tom Reilly, ML engineer at BigCo, asked which pull requests were open on the team's model. Memtro listed them with author, comment count and when opened, and read the timeline of the oldest.
At Acme Ltd, a job starts when a pull request is opened on the company's dataset. It reads the pull request and its commits, then gives Maya Okafor, the data lead, a summary of what it changes before she reviews it.
The teams that lean on Hugging Face most, a question each of them asks, and what they connect beside it.
A record in Hugging Face is rarely the whole story. Connect the systems that hold the rest and Memtro checks them together, then remembers what it learned so the next question starts further along.
Using something else for the same job, or both? Each connector has its own tools; connect as many as you like.
On huggingface.co, open Settings, Access Tokens and create a token. Choose Read for reads, Write to let Memtro act, or Fine-grained to limit it to some repos.
In Memtro, open Connections → Hugging Face and paste the Access token. Set the Namespace to an organization if its Jobs and events should be watched.
An administrator grants the actions by name, such as merging pull requests, to the teams that may use them. Events are turned on from the connection's Events tab.
Ask which pull requests are open on your model.
Chat in the dashboard, the OpenAI-compatible API, scheduled jobs, and any MCP client: Claude Code, OpenCode, Cursor, Claude Desktop. The Hugging Face tools are the same everywhere.
Only when you ask, and only what it has been granted: open a discussion or pull request on a repo; comment on a discussion or pull request; close or reopen a discussion or pull request; rename a discussion or pull request; merge a pull request; create a model or dataset repo; start a Job on paid hardware; cancel a Job; change a Space's hardware; accept or reject a person's access request to a gated repo. Every other tool reads, searches or lists. The Connections page labels the tools that act, an administrator grants each one by name to the teams that may use it, and a personal connection acts only after its owner allows it.
A personal connection serves only your own requests. A shared connection serves the organisation: everyone, or only the teams an admin chooses. Admins can always add shared connections; other members need the permission.
Credentials are encrypted at rest. Tool results are cached briefly so agent loops do not repeat identical calls, and the facts the model extracts are saved to your organisation's memory with provenance (system, record and link) so they can be checked and superseded.
Access token. On huggingface.co, open Settings, Access Tokens and create a token. Choose Read for reads, Write to let Memtro act, or Fine-grained to limit it to some repos.
Yes. Turn events on in the connection's Events tab on the Connections page, choose the kinds you want, and Memtro registers its address with Hugging Face for you. A job set to start "When an event arrives" then runs once for each matching event, with the event as its input. The run treats the event as outside data, never as instructions, and anything the run remembers is kept as unverified.
Seven-day free trial, then one monthly price per organisation. Connectors, jobs and memory are never metered.
Sign in, add a provider key, create an API key and point your tools at Memtro. It takes about five minutes.
7-day free trial · card on file, nothing charged until it ends · cancel any time