From the workshop
Developer Updates
Follow the work as it becomes usable.
The product story now includes a structured workspace, agent and model controls, scoped memory, governed skills and tools, and a dedicated Media Studio interface.

Workspace and continuity
Organizer, projects, folders, conversations, and contextual controls give the application a clearer structure for ongoing work.
Explore this topic →Capabilities and control
Agent management and the Skills & Tools panel make roles and governed actions more visible.
Explore this topic →Knowledge and creative work
Memory Notes and document workflows address context; Media Studio gives configured creative workflows a dedicated interface.
Explore this topic →Release information
This overview is not a release certification or a dated changelog. Availability depends on the deployed version and configuration. Contact the project for the release appropriate to your use.
Explore this topic →September 2026 documentation update
This website refresh adds practical guidance for in-app Help, Organizer navigation, requested-versus-active model attribution, Organization memory scope, editing Memory Notes, and Skills & Tools. These are descriptions of the inspected application source, not a promise that every capability is enabled in every early-access installation.
Ask Friday to guide you
In an authorized FridayLocalAI conversation, try:
Help me understand the FridayLocalAI features available to my role. Use relevant Help topics where available and ask me to confirm the visible controls. Do not infer that a feature is enabled from its name alone.
Tell Friday what you can see and which step you have reached. If the instructions do not match your screen, say so; ask for the relevant Help topic or contact your operator. Friday cannot be assumed to see your screen or know your current settings.
Read this page as a current product overview
This page explains the areas covered by the current documentation refresh. It is not a substitute for release-specific instructions or an acceptance record for every installation. The content is based on inspected application source and the project’s stated availability: early access by contact, no public installer, and no purchases through the FridayLocalAI Store. That distinction matters because a user interface, a configured workflow, and a verified result are separate forms of evidence. Documentation can accurately describe a control without claiming that every user has permission to use it. When planning work, check the features available in your deployment and ask the operator about any capability whose setup or status is unclear.
Workspace guidance is more explicit
The website now explains how Organizer, conversations, and Dashboard support continuing work. Guides describe finding the intended context, using meaningful conversation names, searching for existing work, and checking saved history. They also explain why organization metadata and access permissions are different concerns. The practical goal is to make the first session less dependent on guessing. A new participant can begin with a small practice task and ask Friday for one step at a time. If the interface differs from the instructions, the participant can report the visible labels and resolve the mismatch before proceeding. This is a documentation improvement, not a claim that an unseen workspace action was performed on the user’s behalf.
In-app Help has a clearer place in the workflow
The inspected application includes information icons and a mechanism that can supply relevant active Help topics to suitable conversation questions. The updated guides explain how to use those resources together: read the section guidance, ask a focused question, and compare the answer with the controls and results in the current session. Help content provides contextual explanation. It does not grant a capability, establish that a service is running, or prove that a save completed. The revised wording keeps those boundaries visible while encouraging users to ask for guidance. If a topic is missing or does not answer the question, that is useful feedback for the project rather than a reason to assume the application can infer every detail.
Model and memory descriptions are more precise
The model guide now distinguishes the requested Model selector from active attribution on the response. Agent and capability routing can affect the model used, so comparisons should use actual attribution and consistent test inputs. Earlier replies do not change simply because a future requested model changes. Memory guidance now includes Organization alongside Global, Project, Folder, and Conversation scopes. It describes saving and inspecting a note, selecting an appropriate category, and editing existing notes where permitted. These explanations help users distinguish a durable record from a conversational promise. The application’s context mechanisms support continuity, but they are not automatic training or unrestricted access to every source ever supplied.

Source review and media guidance focus on evidence
The document pages explain the difference between upload acceptance and readable task context. They recommend a narrow, checkable question before a broad analysis and ask users to compare important passages with the original source. This helps identify whether a problem lies in source availability, task scope, or the generated interpretation. Media Studio guidance similarly separates a planned brief, a submitted job, a completed artifact, and a reviewed result. Users are encouraged to inspect the actual output and verify the enabled workflow rather than assuming all media capabilities are available. The documentation does not promise a particular provider, performance level, or offline result for every deployment.
What to watch for next
For release availability and deployment-specific changes, contact the project or follow the instructions supplied by your operator. Do not infer a release date from the website’s theme year or a documentation review date. When an update affects your work, ask which behavior changed, what configuration applies, and what small verification task demonstrates the intended result. You can contribute by reporting a concrete discrepancy between the guide and the application. Include the feature, visible label, expected behavior, and observed result without private contents or credentials. A useful update process connects documentation, implementation, and verification. It avoids treating a broad list of feature names as evidence that every workflow is complete, enabled, and suitable for every participant.
A useful final check
Documentation dates identify a content review. They should not be interpreted as a new software version, a release certification, or a deadline for a capability that has not been announced.