Turn Research into a Weekly Plan

What you get

A dated weekly plan with prioritized tasks, owners, deliverables, and checkpoints, synthesized from existing research.

Who this is for

This scenario is for the weekly ritual: collect, synthesize, plan, execute. The agent handles the synthesis and formatting, while a human supplies the goal, the constraints, and the final priority call.

How the phases connect

Collect checks the material for unreliable sources. Plan turns the goal and the material into a short task list. Produce formats the plan so it reads like a schedule, not a dump. Check verifies the plan fits the time budget and that every task is startable.

Why the plan stays short

A plan that fits on one page gets executed. The workflow deliberately caps scope: three outcomes, one checkpoint, and tasks that start with a verb. Anything that does not fit is parked, not deleted, so the plan stays honest about what one week can hold.

Prerequisites

A set of research notes, bookmarks, or saved articles
A stated goal for the week
An honest estimate of available time

Skill tree

The phases read left to right on desktop and top to bottom on mobile. Capability nodes are tool-neutral; the skills under them are the concrete options.

Goal

Collect

Capability

Fact Verification

Check claims, sources, dates, and links before content is published.

Fact Verification SEO Content Workflow

Plan

Capability

Requirement Planning

Turn a fuzzy goal into a concrete set of requirements the agent can act on.

Codex Astro MVP Research to Plan

Produce

Capability

Writing

Produce clear, useful prose that matches the brief and the audience.

Research to Plan SEO Content Workflow

Check

Capability

Validation

Judge whether the result actually meets the acceptance criteria.

SEO Content Workflow Skill Selection Validation Checklist

Capability breakdown

Each node states why it exists, what it takes in, what it produces, what can go wrong, and how it is accepted.

Fact Verification

AI-generated content fails on confidence, not effort. Unverified claims erode trust and can cause real damage, so every load-bearing fact needs a source.

Input

-Draft claims and quoted figures
-Candidate source URLs

Output

-Verified claim list with source links and check dates
-Unverifiable claims flagged for removal
-Source quality notes

Risks

-Confusing "the model said so" with "a source says so"
-Citing the same low-quality source repeatedly
-Letting verification become an afterthought

Acceptance

-Every factual claim traces to a named, checked source
-Claims without sources are either removed or marked unverified
-Check date is stored with each verified claim

Requirement Planning

Agents fail when the goal is vague. Requirement planning converts "make me a site" into scope, constraints, target users, and acceptance criteria before any code is written.

Input

-A one-line goal or problem statement
-Known constraints (budget, time, platform, audience)
-Existing assets or content

Output

-Scoped requirement list with priorities
-Out-of-scope list
-Success criteria the build can be measured against

Risks

-Scope creep when requirements are written as feature wish lists
-Confusing user wants with user needs
-Skipping the step entirely and letting the agent improvise

Acceptance

-Every requirement is testable and either in scope or explicitly out of scope
-The team agrees on what "done" means before execution starts

Writing

Content is the product for many scenarios. Writing turns research into readable output that answers the user's question without filler.

Input

-Outline and source material
-Style and tone guidance
-Target reader description

Output

-Draft sections or full article
-Internal links and metadata suggestions
-Ready-for-review copy

Risks

-Uniform AI voice across every piece
-Padding length instead of adding information
-Ignoring the reader's actual question

Acceptance

-The piece answers the stated question in the first third
-No unverified claims survive the draft
-Tone and structure match the brief

Validation

The final step decides whether work is done. Validation turns "looks finished" into a checklist of observable, repeatable checks.

Input

-Acceptance criteria from planning
-The delivered output or site

Output

-Completed checklist with pass/fail per item
-Evidence (screenshots, logs, URLs)
-Fix list for failed items

Risks

-Subjective "good enough" instead of explicit criteria
-Validating only what is easy to check
-Marking items passed without evidence

Acceptance

-Every criterion is a yes/no check, not a feeling
-Failed items produce a concrete fix and a re-check
-Validation is repeatable by a different agent

Recommended build

A ready combination of skills for this scenario, with the order and checks included.

Minimal build

Research to Plan Starter

A short weekly ritual that turns saved research into a one-page plan with owners and checkpoints.

Skills: 3 skills
Steps: 6 steps

Execution order

The recommended sequence from the Research to Plan Starter build.

1
Review the material and flag unreliable sources
2
State the goal and the time budget
3
Draft the plan capped at three outcomes
4
Write tasks with verb, deliverable, and owner
5
Add a mid-week checkpoint
6
Run the plan checklist and adjust

Open the full build

Validation checklist

The plan fits the stated time budget
Every task has a verb, a deliverable, and an owner
Priorities are explicit and justified
A mid-week checkpoint is scheduled
Tasks are concrete enough to start without more research

Common failures

The plan lists everything instead of priorities

Cap the plan at three outcomes and rank them

Tasks are too vague to start

Rewrite each task with a verb and a deliverable

The plan ignores the actual time available

State the time budget before planning and audit the result

Give this brief to your agent

Copy the structured brief, adapt the out-of-scope list, and ask the agent to return the checklist with evidence.

Agent task brief

Goal: A dated weekly plan with prioritized tasks, owners, deliverables, and checkpoints, synthesized from existing research.
Scenario: Turn Research into a Weekly Plan
Categories: Research, Automation
Difficulty: beginner

Prerequisites:
- A set of research notes, bookmarks, or saved articles
- A stated goal for the week
- An honest estimate of available time

Capabilities in execution order:
Phase: Collect
- Fact Verification
Phase: Plan
- Requirement Planning
Phase: Produce
- Writing
Phase: Check
- Validation

Recommended build: Research to Plan Starter
Execution steps:
1. Review the material and flag unreliable sources
2. State the goal and the time budget
3. Draft the plan capped at three outcomes
4. Write tasks with verb, deliverable, and owner
5. Add a mid-week checkpoint
6. Run the plan checklist and adjust

Validation checklist:
- [ ] The plan fits the stated time budget
- [ ] Every task has a verb, a deliverable, and an owner
- [ ] Priorities are explicit and justified
- [ ] A mid-week checkpoint is scheduled
- [ ] Tasks are concrete enough to start without more research

Out of scope:
- Research that has not been collected yet
- Plans that require deep domain context the agent lacks

Evidence required:
- Checklist with pass/fail per item
- Screenshots at the agreed viewports
- Build or run log from a clean checkout