dcp-lead-init

name	dcp-lead-init
description	Start Deep Co-Planning as the lead for an active planning run.
invocation-args	{"task_name":"required"}

dcp-lead-init

Do not start this playbook implicitly.

Bind the lead run to an explicit task_name at invocation time.

Read First

1. .agent/playbooks/deep-co-planning/PLAYBOOK.md
2. .agent/playbooks/deep-co-planning/transport.md
3. .agent/playbooks/deep-co-planning/template/round_template/README.md

Keep these hot-loaded for the duration of the playbook:

• .agent/playbooks/deep-co-planning/PLAYBOOK.md
• .agent/playbooks/deep-co-planning/skills/dcp-lead-init/SKILL.md

Your Job

You are responsible for:

• initializing the planning workspace with human guidance
• co-authoring the HLD with the human
• maintaining the live round packet
• ingesting reviewer outputs
• classifying reviewer feedback
• producing reviewer-facing carry-forward feedback
• producing human-facing round results
• deriving one or more specs from the approved HLD
• maintaining known unknowns inside the HLD or final specs
• closing each round with the human's verdict

Operator Path

Use the CLI to initialize the runtime:

dotagent playbook init deep-co-planning --task <task_name>

Do not manually scaffold the planning filesystem unless you are explicitly handling a fallback case.

Only the lead updates shared synthesis artifacts.

Round Creation Rules

• initialize the first round through the CLI
• create another round only when the human explicitly asks for one
• when another_round is requested, create the next round, carry forward the required artifacts, and wait until the new packet is ready before reviewers begin
• do not pre-create future rounds
• do not ask reviewers to start until the active round packet is ready

Before Reviewers Start

Confirm:

• planning objective
• implementation-phase scope
• whether the active round is hld-review or spec-review
• the exact artifacts under review
• reviewer roster
• reviewer domain assignments when present
• approval target for the round
• carry-forward artifacts when this is not the first round

If those are incomplete, the round is not ready.

For every new round, start from the active round README.md.

Treat 00_round_context.md as the full round contract.

Human Interaction Points

Engage the human when:

• the run is just starting and the planning target still needs to be understood through open discussion
• scope is too broad or too narrow for this playbook
• the HLD needs correction before reviewers start
• reviewer domain assignment is unclear
• open questions arise during a round
• an approved verdict is being requested for the HLD or the final specs
• another round is recommended

At kickoff, start naturally.

Do not rush straight into HLD prompts.

Prefer opening questions such as:

• what do you want to build?
• what do you want to plan in this session?

Use the discussion to discover the real planning target, then translate it into the HLD structure.