agent-task-splitter

Turn one approved goal into a provider-neutral role DAG and bounded task packets. This skill plans; it does not spawn agents.

Read references/tasksplitterheuristics.md when role selection or DAG shape is not obvious. Read ../../docs/public-harness-contract.md for schemas and artifact policy.

Roles

• primary-agent: owns scope, architecture, and human communication.
• delegated-executor: performs bounded implementation or mechanical work.
• reviewer: independently tests and judges a stable candidate.
• synthesizer: structures completed inputs without reopening discovery.

The host chooses the adapter for each role. Do not put provider or model names in the public role field.

Use this skill when

• Two or more independent task packets can run in parallel.
• Implementation and independent review must be separate.
• A fan-out/fan-in or diamond DAG materially shortens the critical path.
• Several results require explicit reconciliation and acceptance.

Do not split exploratory debugging before the cause is known. Do not split a small coherent change merely to create agent activity.

Inputs

• Goal and authorized scope.
• Success criteria.
• Files, systems, and external actions in/out of scope.
• Available policyref and checkpointref when any child may be spawned.
• Existing recorded human decisions and evidence refs.

If scope or acceptance is ambiguous, ask one focused question. Do not invent permission for external writes.

Workflow

1. Confirm the repository/worktree root.
2. Restate the goal and scope.
3. Identify task boundaries by work character and evidence dependencies.
4. Assign one role to each task.
5. Build an acyclic dependency graph.
6. Partition write scope. Two parallel writers must not own the same file.
7. Add at least one runnable or objectively checkable success criterion per

task.

1. Write .coord/plan.yml using schema_version 2.
2. Write .ai/task<NNN><slug>.md for each non-inline task.
3. Return the ready task ids and dependency order. Do not spawn.

Plan shape

schemaversion: 2 round: 1 goal: "..." policyref: "${AGENTCOLLABPOLICY}" checkpointref: ".coord/task-checkpoint.json" createdat: "<ISO 8601 with timezone>" tasks: - id: T1 role: primary-agent slug: define-contract description: "Freeze the public contract." dependson: [] filesinscope: ["docs/contract.md"] filesoutofscope: ["src/"] successcriteria: - "contract is traceable to the current authorized user goal" - id: T2 role: delegated-executor slug: implement-contract description: "Implement the approved contract." dependson: [T1] filesinscope: ["src/", "tests/"] filesoutofscope: ["docs/contract.md"] successcriteria: - "python -m pytest tests -q" - id: T3 role: reviewer slug: review-candidate description: "Review the stable T2 candidate." dependson: [T2] filesinscope: [] filesoutofscope: ["/*"] successcriteria: - "verdict is PASS, FAIL, or NEEDSHUMAN with evidence"

Omit policyref/checkpointref only when the plan cannot spawn or loop autonomously.

Task packet

# Task: <id> — <description>

## Context - Repo/worktree: <absolute path> - Plan: .coord/plan.yml - Role: <role> - Depends on: <task ids and artifact refs>

## Pre-task scope confirmation Before editing, report the exact allowed and forbidden paths. Stop if the brief conflicts with the plan.

## Goal <one bounded deliverable>

## Scope - May read: <paths> - May write: <paths> - Must not touch: <paths> - External actions: <none or explicit authorization>

## Acceptance - <runnable or objective criterion>

## Return contract - status - concise summary - fileschanged - testsrun - evidence_refs - risks - blockers

Task packets and raw results are scratch. A task may write only its explicit shipping artifact; acceptance evidence is promoted separately.

Policy boundary

Immediately before a host spawns a task:

agent-collab policy evaluate \ --policy <policyref> \ --checkpoint <checkpointref> \ --json

The host must not spawn unless decision=continue and spawn_allowed=true. Splitter output does not override that decision.

The host may delegate read-only exploration while planning, but must not turn a planning-only request into implementation. Prefer direct execution for small coherent work. Reserve capacity for required independent review and retain cumulative child usage when a v2 slice advances. Agent boundaries alone do not require commits or fresh human authorization.

Invariants

• Provider names are transport metadata, not public roles.
• Reviewer and synthesizer are different: a synthesizer structures accepted

inputs; a reviewer judges them.

• A task cannot approve its own semantic or governance-sensitive output.
• Missing, null, failed, declined, cancelled, and timed-out tasks stay

non-success.

• Every parallel result list filters absent/failed results before downstream

synthesis while retaining their failure records.

• Agent voting never replaces evidence verification or a human gate.

Compatibility

Historical schema-less plans and provider-specific task paths are parse-only. Writers emit v2 roles and generic task paths. See ../../docs/migration-0.4.md.