FAQ
Answers to common questions about how Giga works.
This FAQ covers the current Giga product model.
For live workspace state such as current permissions, connected tools, available models, schedule slots, or surface-specific capabilities, use the current product as the source of truth.
What is Giga?
Giga is a permission-aware AI workspace built around company context, ongoing work, reusable AI roles and methods, connected systems, and recurring processes.
Its core primitives include Chats, Groups, Tracks, Context Notes, Agents, Skills, Workflows, Pulses, Connections, files, and databases.
What is Company Brain?
Company Brain is Giga’s permission-aware company context layer.
Depending on relevance and access, it can draw from workspace and Group guidance, Tracks, Context Notes, Observations, Current Track State, files, Connections, Agents, Skills, Workflows, and user-level context.
The current user and workspace permissions determine what Giga can retrieve and use.
What is the difference between a Chat and a Track?
A Chat is the direct working surface for a task or discussion.
A Track is the durable home for work that continues over time.
Use a Track when decisions, files, memory, or progress need to accumulate across multiple Chats.
A Chat can link to at most 1 explicitly selected Track.
Can Giga choose a Track automatically?
Track attachment should be explicitly selected or confirmed by the user.
That matters because Track attachment affects where Observations from the Chat can become retrievable.
Giga can help identify relevant Tracks, while the actual attachment stays explicit.
What is a Context Note?
A Context Note is the durable standing context for a Track.
Use it for stable background, terminology, constraints, standing instructions, and durable handoff context.
Changing progress and decisions belong in Track memory or Current Track State.
What is Current Track State?
Current Track State is the curated current position and decision record for a Track.
It captures where the work stands, how it got there, what was ruled out and why, and what remains open.
What is the difference between an Agent and a Skill?
An Agent is a named reusable AI role with instructions, context, Skills, and permitted Connections.
A Skill is a reusable task method or instruction set.
Use an Agent when the reusable asset is the role. Use a Skill when the reusable asset is how a task should be done.
What is the difference between a Skill and a Workflow?
A Skill stores a reusable method for a task.
A Workflow stores a reusable ordered process with several meaningful stages.
For example, a review standard fits a Skill. A process that gathers information, evaluates it, and produces a handoff fits a Workflow.
What is the difference between a Workflow and a Pulse?
A Workflow defines a reusable process.
A Pulse defines recurring scheduled work.
A Pulse can target an Agent, target a Track, or run without a target.
Any exact relationship between Pulses and Workflows should follow the current live product capabilities.
What is a Group?
A Group is a placement and permission scope.
Groups can scope Tracks, Agents, files, credentials, Connections, and shared context.
Use a Group when a real team, shared area, or access boundary needs its own place.
Can I keep work private?
Each workspace member has an existing private Group for private work.
A private Track should live in that existing private Group.
Creating another Group purely to make a single Track private is generally unnecessary.
Learn about Personal vs shared work →
Does being in the workspace give me access to everything?
Access is layered.
Workspace membership, workspace role, Group membership, resource ownership, Connection sharing, provider permissions, and the current live tool surface can all affect what someone can use.
Can Connections be shared?
A credential can be owned by one user and shared into Groups.
A teammate may be able to use a shared Connection while management actions such as renaming or deleting the credential remain with the owner.
The external provider’s own permissions still apply.
Learn about Sharing Connections →
Can Giga connect to custom tools?
Current custom Connection paths can include:
- MCP
- OAuth with supported self-serve registration
- API key or PAT
- browser login
The external service and current Giga integration surface determine which method is available.
Learn about Custom Connections →
Should I paste API keys or passwords into Chat?
Use the secure connection flow for authentication material whenever it is available.
Secrets such as API keys, passwords, tokens, OAuth secrets, and connection strings belong in the secure credential layer rather than ordinary Chats, files, or Context Notes.
Can Giga use websites when there is no API?
Browser-based work can be used when a stronger first-class route such as an API, MCP tool, database, or built-in integration cannot complete the task.
For logged-in work, Giga can use the current browser capability where available.
CAPTCHA, bot detection, or two-factor authentication can require the user to complete the login step.
Learn about Browser Connections →
Can Giga create files?
Giga supports authored work product through files.
Confirmed capabilities include Markdown, text and binary Track files, code, webpages, and visual artifacts where the current surface supports them.
SQLite and hosted Giga databases can be used for structured growing records.
Exact rich file formats should be verified from the current product capability.
Learn about Working with files →
Can I use Giga in Slack, ChatGPT, Claude, or Microsoft Teams?
Yes. Giga’s official user-facing surfaces are:
- Giga Web
- Slack
- ChatGPT
- Claude
- Microsoft Teams
The underlying Giga workspace context and permission model stays consistent, while file behavior, UI, and exact available capabilities can differ by surface.
Learn about using Giga anywhere →
Does Giga support different AI models?
Model choice and provider flexibility are a first-class Giga product area.
Exact providers, model availability, plan access, custom endpoints, enterprise agreements, and bring-your-own-model options should be verified from the current product configuration.
Giga’s workspace context and permissions remain conceptually separate from any single model provider.
Can a Pulse run without an Agent?
Yes.
A Pulse can currently target:
- an Agent
- a Track
- no target
The live Pulse surface determines the exact schedule options and targets available.
What schedules do Pulses support?
The current Pulse model supports:
- shared named schedule slots
- custom schedules
- daily cadence
- weekly cadence
- monthly cadence
- intra-day cadence
What happens when a workspace member is removed?
The current member-removal operation soft-archives the membership and preserves the member’s data.
Owners cannot be removed through the current removal operation.
Removing a member is a consequential access change and should be confirmed.
Where should I put company-wide instructions?
Use workspace context for durable guidance that genuinely applies across the company.
Use Group context for team-specific guidance.
Use a Context Note for standing guidance tied to one Track.
Learn about Workspace settings →
I still do not know which feature to use
Start with the job:
- direct task → Chat
- team or access boundary → Group
- ongoing work → Track
- stable Track background → Context Note
- reusable AI role → Agent
- reusable task method → Skill
- reusable ordered process → Workflow
- recurring scheduled work → Pulse
- live external system → Connection
See Choosing the right Giga feature →