Profiles, Goals, And Agents

These three features work together but solve different problems. An Agent Profile is a reusable role and policy. A goal is a bounded objective for the current task. An agent run is one execution of delegated work, with its own status and evidence.

For the mental model behind delegation and hand-offs, read How Profiles, Goals, And Agents Work.

Agent Profiles package reusable instructions, skills, tools, model policy, and approval policy without changing the underlying thread history.

Choose Or Create A Profile

1. Open a chat and expand Agent profiles in the left sidebar.
2. Choose a built-in profile or create one for a repeated role.
3. Keep the role narrow: describe the outcome, sources it may use, and what it must not do.
4. Review its model, skill, tool, delegation, and approval policy.
5. Save it, then select it before asking for work.

Profiles do not grant capabilities by themselves. A selected tool must still be enabled, available, and allowed by the active approval policy. A profile that can delegate may start child agents, but each run remains visible in the Activity Center and the chat agent strip.

New runs also receive a checkpointed work budget. The Models settings tab shows the recommended application-wide work-round, nesting, concurrency, and optional child active-time limits. Extra children wait in a first-in, first-out queue when capacity is full; changing a setting affects new runs, not work already in progress.

The goal strip records bounded progress while parent and child agent rows show who is running, complete, waiting, or stopped.

Run A Goal

A goal is useful when work needs more than one turn. State the objective and an observable finish condition. Row-Bot records progress, blockers, evidence, turns, and status so it can continue without pretending that partial work is complete.

• Active means work can continue.
• Complete means the objective and required checks are genuinely finished.
• Blocked is for a repeated impasse that requires your input or an external change.
• Stop ends the active run; it does not delete the thread or its evidence.

Review Delegated Work

Parent agents coordinate. Child agents handle bounded subtasks. Open an agent row to review its prompt, profile, model, status, result, evidence, and thread. A completed child result is evidence for the parent, not automatic permission to write files, send messages, commit code, or publish.

For coding work, a parent can assign a child to an existing local folder. Row-Bot registers that folder as the child's Developer workspace and applies writer ownership to the resolved folder, not merely the process working directory. Children assigned to distinct folders or worktrees can write in parallel; writers targeting the same folder wait and run serially.

If a run appears stuck, inspect its last update before stopping it. If several agents edit the same resource, give them distinct folders or worktrees, narrow their ownership, or run them sequentially. Keep consequential final actions with the parent and a human approval point.

Row-Bot saves the parent's first continuation checkpoint when asynchronous delegation begins. If a child process outlives its launching shell or the app restarts, resume repairs only genuinely unanswered tool calls with an interruption marker, then wakes the parent from durable child results. It does not replay previously executed work. If a checkpoint cannot be repaired safely, the run remains stopped for review.

Repeated model-and-tool states are detected before a run can loop indefinitely. The fourth identical no-progress state is blocked; a fifth ends the run cleanly with a durable reason. Reaching a configured work limit also finalizes the run instead of leaving it marked as active.

What Is Saved

Profiles, goals, agent runs, edges, progress, results, and evidence are stored in the active local data directory. Provider prompts still follow the privacy terms of the model route selected for each run.