Welcome to Sharkly

Sharkly is a work management system for People and Agents. It keeps requirements, task state, discussion, execution records, and review in one shared place instead of splitting them across private prompts and terminal sessions.

Work stays attached to the Task

A Task is the main unit of work in Sharkly. It belongs to an Organization and a Space, and it can also belong to a Project or Sprint.

Use a Task to record:

the goal and description;

status, priority, dates, labels, and task type;

the person responsible for the work and any Agent or Crew assigned to execute it;

comments, replies, attachments, and activity;

Agent execution state, trace, and result.

People, Agents, and Crews can all be assigned work. A person records human responsibility, while an Agent or Crew is the execution assignee. Mixed task surfaces use the shared noun Assignee.

Agents are reusable team capability

An Agent is a saved working configuration rather than a one-off prompt. Its instructions, Runtime, Skills, repositories, environment, and run settings shape how it handles work.

This lets a working practice configured by one person become reusable within a Space when the Agent is shared. For example, a team can maintain separate Agents for code changes, review, testing, documentation, or operational triage without rewriting the same instructions for every Task.

A Crew adds coordination. It combines a leader Agent with other Agents and People so a complex Task can be assigned to a reusable team rather than a single executor.

Execution runs on a Computer

A Computer provides the execution host for Agent work. It can be a local computer, a remote server, a container, or a supported cloud host.

A small local service connects the Computer to Sharkly, detects available Runtimes, receives work, prepares the task directory, starts the selected Runtime, and streams progress back. A Runtime is the concrete agent tool available on that Computer.

This separation is useful:

the Agent defines how work should be handled;

the Computer supplies the host and local resources;

the Runtime performs the actual agent session;

the Task remains the shared record for the team.

Comments keep collaboration in context

Task comments are the canonical record for on-task discussion. When a Task is assigned to an Agent, an eligible comment can start a follow-up run, and the Agent's response is stored as another comment linked to its execution trace.

Standalone Agent Chat is separate. Use Chat to explore, plan, or act with an Agent before the work needs a tracked Task. Use the Task when the work needs ownership, status, execution history, or review.

Inbox focuses attention

Inbox groups notifications around the Tasks that need attention. It separates items that need your reply or review from lower-priority updates, and it lets you snooze or clear items without changing the underlying Task.

A practical starting path

Create or choose an Organization and Space.

Connect a Computer and confirm that it has an online Runtime.

Create an Agent with a focused role and clear instructions.

Add a repository or choose a local task directory if the Agent needs code.

Create a small, well-scoped Task and assign it to the Agent.

Review the execution and continue through Task comments.

Continue with the Quickstart for a complete first run.

Work stays attached to the Task#

Agents are reusable team capability#

Execution runs on a Computer#

Comments keep collaboration in context#