{#- The Plan tab: the task's plan with its checkpoints, the bound agent's todo (a second layer, never merged into the plan) and the plan's history. Expects: t (the task), plan, todo, history, optional error. Refetched on live events. -#} {%- macro todo_items(items) -%}
Now doing: {{ now.active_form or now.subject }}
{% endif %}{{ error }}
{% endif %} {% if plan.steps %}Added after acceptance by {{ s.added_by }} : {{ s.added_reason }}
{% endif %} {% if s.details_md %}{{ s.conformance or "done" }} {%- if s.outcome %} · {{ s.outcome }}{% endif %} {%- if s.link %}{% set href = s.link|http_url %} · {% if href %}{{ href }}{% else %}report #{{ s.link[7:] }}{% endif %}{% endif %} {%- if s.done_at %} · {% endif %}
{% elif s.outcome %}{{ s.outcome }}
{% endif %} {% if is_current and todo %}{{ todo_block(todo, false) }}{% endif %}No plan yet. The agent writes 2-5 steps with plan_set and marks them with plan_step_update; a small (S) task may go with the agent's todo only.
{% for line in diff %}{{ line }}
{% endfor %}{% endif %}