How to Plan a Month of Content with Claude + Notion
Turn real audience questions, source material, and available time into a 30-day editorial plan. The workflow below instructs Claude to preview the month, show its evidence and workload, and ask for approval before anything reaches Notion.
Start with constraints, not 30 blank ideas.
Give Claude one audience, one monthly goal, your usable source material, production capacity, formats, and blackout dates. Ask for a preview that maps pillars, anchors, derivatives, review days, and buffers. Check workload and evidence, then approve the exact Notion destination and property mapping. Without a connector, export a UTF-8 CSV and import it manually. This creates an editorial calendar; it does not publish posts.
01 · INPUT BRIEF
Start with the inputs Claude cannot infer
A polished calendar can still be fiction. Claude cannot know your real production time, audience language, product facts, or blackout dates unless you provide them. Collect these six input groups before asking for ideas. If a required field is missing, the prompt later in this guide tells Claude to stop and ask.
Window and timezone
Start date, end date, timezone, blackout dates, launch dates, and any review or approval lead time.
One audience
Who this month is for, what they are trying to do, and questions you have actually heard from them.
One monthly goal
The business or creator outcome for the month, plus the one next action each content cluster may support.
Usable source material
Notes, transcripts, comments, customer questions, product facts, existing posts, and links Claude may cite instead of inventing evidence.
Channels and formats
Primary platform, secondary channels, formats you can make, and platform-specific constraints you already know.
Production capacity
Hours available, realistic effort by format, batching opportunities, collaborators, and a buffer for work that slips.
02 · DESTINATION
Choose how the calendar will reach Notion
Pick the delivery path before planning so the output uses fields your destination can accept. Anthropic describes a Notion connector that can work with content available through the authorized connection. Notion also provides a hosted MCP option for supported AI clients. Both require connection and authorization. That authorization can let the connected client read or update content your user can access in the selected workspace. The destination preview in this guide is a Kabo operating rule—not a page-level permission boundary.
Anthropic: Notion connector for ClaudeNotion Developers: connect an AI tool to Notion MCPNotion Developers: MCP security best practices
| Path | Use it when | Keep this boundary |
|---|---|---|
| Claude + Notion connector | Your Claude client exposes the connector and you have authorized the selected Notion workspace. | Preview the database, property mapping, and rows; approve only that proposed write. |
| Notion hosted MCP | Your AI client supports MCP and you have completed the Notion OAuth flow. | Confirm the active workspace and destination before any create or update action. |
| Markdown + UTF-8 CSV | No connector is available, or you want a file you can inspect and import yourself. | A CSV import adds rows; it is not sync, upsert, social scheduling, or publishing. |
Notion Help: import data into Notion
Use a small schema you can audit
Start with the properties below and add complexity only when a real workflow requires it. A Notion calendar view needs a date property. A content row being visible in that view does not mean the post is scheduled on a social platform.
Notion Help: calendar viewsNotion Developers: tools supported by Notion MCP
| Property | Suggested type | Why it exists |
|---|---|---|
| Publish date | Date | Drives the calendar view |
| Working title | Title | Names the content row |
| Platform | Select | Names the intended destination |
| Format | Select | Reel, carousel, email, story, live, or another real format |
| Pillar | Select | Keeps the month balanced around a small set of themes |
| Content role | Select | Anchor, derivative, launch, community, review, or buffer |
| Source / evidence | Rich text; use URL when every value is a link | Shows what the claim or idea came from |
| CTA | Text | Records one intended next action, including none |
| Effort hours | Number | Lets the month be checked against capacity |
| Status | Select | Idea, Planned, Approved, Drafting, Ready, Scheduled, or Published |
03 · MONTH ARCHITECTURE
Build anchors before you fill 30 dates
Generating one isolated idea for every day encourages repetition disguised by different titles. Instead, build a few evidence-backed anchors: a useful tutorial, point of view, workshop, experiment, or launch moment. Then derive narrower pieces that answer a different question or use a different depth and format.
- 01
Choose two to four pillars
Each pillar should connect the monthly goal to source material you already have. Drop a pillar that is supported only by Claude's general knowledge.
- 02
Give every anchor a source and payoff
Record the notes, transcript, question, example, or product fact behind it. Name what the audience can do or understand after consuming it.
- 03
Make derivatives earn their own row
A derivative should change the question, evidence, depth, or action. Cropping the same claim into three aspect ratios is distribution work, not three ideas.
- 04
Reserve review and buffer time
Add production estimates before dates are locked. Kabo's working rule is planned hours plus a buffer must fit inside available hours. The prompt defaults to a 20% buffer, but you can replace it with your own known margin.
If an anchor needs a stronger opening, use the observable patterns in Kabo's guide to what viral videos have in common. Treat those patterns as editing prompts—not a promise that the finished post will go viral.
04 · COPYABLE PROMPT
Use this prompt to build the month
Replace every bracketed field. The prompt separates planning from writing: it instructs Claude to audit the inputs, calculate the load, show the full month, and report QA issues. It treats a second, exact approval phrase as a review convention for one destination and row set.
You are helping me plan one month of creator content. Build a realistic editorial calendar from the information I provide. Do not invent audience research, analytics, customer quotes, product facts, platform rules, trends, deadlines, or performance predictions. REQUIRED INPUTS - Planning window: [START DATE] to [END DATE] - Timezone: [TIMEZONE] - Audience: [WHO THIS MONTH IS FOR] - Primary platform: [PLATFORM] - Secondary channels: [CHANNELS OR NONE] - Monthly goal: [ONE GOAL] - Allowed CTAs: [CTA OPTIONS, INCLUDING NONE] - Content pillars: [2–4 PILLARS] - Available source material: [NOTES, TRANSCRIPTS, COMMENTS, LINKS, PRODUCT FACTS] - Available production hours: [HOURS] - Effort by format: [FORMAT = HOURS] - Blackout, launch, review, and approval dates: [DATES] - Notion destination: [DATABASE NAME + URL/ID, OR "NO CONNECTOR"] - Existing Notion properties: [PROPERTY NAMES + TYPES, OR "NEW DATABASE"] STOP CONDITIONS If the planning window, timezone, audience, primary platform, monthly goal, or production capacity is missing, stop and ask only for the missing information. If a source is missing for a factual claim, keep Status as "Idea," write "Needs evidence" in Notes, and do not fill the gap with a plausible statement. PLANNING METHOD 1. Restate my constraints in a short input audit. Separate supplied facts from assumptions and unresolved questions. 2. Propose 2–4 content pillars. For each pillar, cite the supplied source material that supports it. 3. Choose a small set of anchor pieces before filling dates. Derive platform-specific follow-ups from each anchor, but give every derivative a different audience question, hook, depth, or next action. 4. Estimate production hours using my format estimates. Keep planned work inside capacity and preserve at least 20% of available time as buffer unless I provide a different rule. 5. Place anchors, derivatives, community posts, review days, launch moments, and buffers in a logical sequence. Respect blackout dates. 6. Use one Status vocabulary only: Idea, Planned, Approved, Drafting, Ready, Scheduled, Published. Put blockers such as missing evidence, pending audience replies, and blackout reasons in Notes rather than inventing extra status values. Do not treat a Notion calendar as a social publishing system. PREVIEW OUTPUT First return: A. Input audit: supplied facts, missing information, and assumptions that need approval. B. Capacity summary: available hours, planned hours, buffer hours, and the calculation. C. Pillar and anchor map: pillar, anchor, derivatives, source/evidence, goal, and CTA. D. A full calendar preview with one row per calendar item; dates may repeat across channels. Use these columns: Publish date | Working title | Platform | Format | Pillar | Content role | Source / evidence | CTA | Effort hours | Status | Notes E. A QA report covering capacity, audience fit, evidence, distinctness, repurpose quality, platform fit, goal/CTA, calendar logic, Notion safety, and state honesty. NOTION WRITE GATE Do not create, edit, or delete anything in Notion during the preview. After I review the preview, show: - the exact workspace/database destination; - existing property -> proposed property mapping; - number of rows to create and rows to update; - duplicate check and date/timezone check; - any unsupported or uncertain field types. Wait for the exact phrase APPROVE NOTION WRITE before making a Notion change. Approval applies only to the destination, mapping, and rows shown in the latest preview. If anything changes, preview it again. NO-CONNECTOR FALLBACK If no Notion connection is available, return the approved calendar as both a Markdown table and UTF-8 CSV using the same columns. In the CSV, format Publish date as MM/DD/YYYY with the full year so Notion can parse it as a date. State clearly that importing a CSV adds database rows; it is not automatic sync, upsert, scheduling, or social publishing.
05 · HUMAN REVIEW GATE
Check the plan before Claude writes to Notion
Read the calendar like an editor and a production manager. A pass means the plan is safe to place in the database—not that every post is finished, scheduled, or likely to perform.
- 01
Capacity
Do planned production hours fit inside available hours after preserving a buffer?
- 02
Audience fit
Can every publishable row name the same intended audience and a real question, need, or tension?
- 03
Evidence
Does every factual or audience claim have a supplied source, or a visible hypothesis label?
- 04
Distinctness
Are the hook, promise, and payoff meaningfully different from nearby rows?
- 05
Repurpose quality
Does each derivative change the angle, depth, or delivery instead of copying the anchor?
- 06
Platform fit
Is the format native to the named platform, with enough context to stand on its own?
- 07
Goal and CTA
Does each CTA support the monthly goal, and are some rows allowed to teach without asking?
- 08
Calendar logic
Do anchors precede derivatives, review days precede deadlines, and blackout days remain protected?
- 09
Notion safety
Have you checked the destination, property mapping, row count, dates, and likely duplicates before writing?
- 10
State honesty
Are planned, approved, scheduled, and published kept separate so a calendar never implies work is live when it is not?
06 · CONSTRUCTED EXAMPLE
See the workflow in a constructed 30-day month
Maya Chen is a fictional freelance-design educator with a fictional 18.4K Instagram audience and 2.7K newsletter list. She is launching a fictional client-onboarding mini-course and has 34 production hours for November 2026. None of the names, audience figures, time estimates, content results, or rows below describe a Kabo customer or provide a benchmark.
11/01/2026–11/30/2026
America/New_York; all CSV dates include the full year.
Instagram + email
Instagram is primary; the newsletter is secondary.
Qualify course interest
Teach first, then invite aligned freelancers to the fictional course during the launch window.
Notes, questions, product facts
Scope-change notes, process notes, audience responses, the workshop outline, and a verified fictional course FAQ.
34 hours
Format estimates are supplied; unallocated time stays free.
Launch, review, blackout
Launch starts 11/22; reviews are weekly; 11/11 and 11/26 are protected.
Available capacity
Her fictional November production budget.
Planned work
Production, community, and review estimates in the table.
Unallocated margin
Time kept for slippage, edits, and work not yet known.
| Date | Platform | Pillar | Role / format | Working idea | Source / evidence | CTA | Hours | State | Notes |
|---|---|---|---|---|---|---|---|---|---|
| 11/01/2026 | Scope control | Anchor · Carousel | The three places freelance design scopes quietly expand | Five anonymized scope-change notes | Save the scope checklist | 2.5 | Planned | — | |
| 11/02/2026 | Scope control | Derivative · Stories | Poll: where did your last project expand? | Nov 1 anchor | Answer the poll | 0.25 | Planned | — | |
| 11/03/2026 | Scope control | Derivative · Reel | The revision request that is actually a new deliverable | Scope-change note 2 | Save for the next proposal | 1.25 | Planned | — | |
| 11/04/2026 | Scope control | Community · Reply post | Answer the strongest question from the scope poll | Audience response required | Reply to the selected question | 0.25 | Planned | Hold until a real audience response exists. | |
| 11/05/2026 | Scope control | Derivative · Newsletter | A five-line scope checkpoint before design starts | Nov 1 anchor + contract notes | Save the five-line checkpoint | 1.5 | Planned | — | |
| 11/06/2026 | Internal | Scope control | Buffer · No publish | Catch up or leave empty | Capacity rule | None | 0 | Planned | Protected production buffer; no publish planned. |
| 11/07/2026 | Internal | Scope control | Review · Planning | Check evidence, comments, and next-week load | Calendar QA | None | 0.5 | Planned | — |
| 11/08/2026 | Sustainable pricing | Anchor · Reel | Why a cheaper project can cost more creative energy | Time logs from three fictionalized project types | Save for the next quote | 2 | Idea | Needs an evidence check before approval. | |
| 11/09/2026 | Sustainable pricing | Derivative · Stories | Show the energy-cost worksheet, not private client data | Nov 8 anchor | Reply with hidden labor | 0.25 | Planned | — | |
| 11/10/2026 | Sustainable pricing | Derivative · Carousel | Four questions to price coordination, not just design time | Nov 8 anchor + own process notes | Save the four questions | 1.5 | Planned | — | |
| 11/11/2026 | Internal | Sustainable pricing | Blackout · No publish | Client delivery day | Supplied blackout date | None | 0 | Planned | Supplied blackout date; no publish planned. |
| 11/12/2026 | Sustainable pricing | Community · Q&A | Answer one pricing-process question with a real example | Audience question required | Ask a pricing question | 0.5 | Planned | Hold until a real audience question exists. | |
| 11/13/2026 | Sustainable pricing | Derivative · Newsletter | Separate creative labor from project-management labor | Nov 8 anchor | Reply with a cost you miss | 1.5 | Planned | — | |
| 11/14/2026 | Internal | Sustainable pricing | Review · Planning | Check the first two clusters for repeated promises | Calendar QA | None | 0.5 | Planned | — |
| 11/15/2026 | Client onboarding | Anchor · Live workshop | Turn a vague inquiry into a usable project brief | Anonymized inquiry patterns + workshop outline | Join the live workshop | 2 | Planned | — | |
| 11/16/2026 | Client onboarding | Derivative · Reel | One question that exposes a missing decision-maker | Nov 15 workshop | Save the question | 1 | Planned | — | |
| 11/17/2026 | Client onboarding | Derivative · Carousel | The six fields in Maya's pre-project brief | Nov 15 workshop | Save the brief fields | 1 | Planned | — | |
| 11/18/2026 | Client onboarding | Community · Stories | Question box for brief-review objections | Nov 15 workshop | Submit a brief question | 0.25 | Planned | — | |
| 11/19/2026 | Internal | Client onboarding | Buffer · No publish | Catch up before launch cluster | Capacity rule | None | 0 | Planned | Protected production buffer; no publish planned. |
| 11/20/2026 | Client onboarding | Derivative · Newsletter | A pre-project brief readers can adapt | Nov 15 workshop | Join the course waitlist | 1.5 | Planned | — | |
| 11/21/2026 | Internal | Client onboarding | Review · Planning | Confirm launch facts, links, and CTA density | Launch checklist | None | 0.5 | Planned | — |
| 11/22/2026 | Course launch | Launch · Reel | What the client onboarding mini-course helps organize | Verified course outline | See the course outline | 1.5 | Planned | — | |
| 11/23/2026 | Course launch | Launch · Carousel | Course module map with an explicit enrollment CTA | Verified course outline | Enroll in the course | 1 | Planned | — | |
| 11/24/2026 | Course launch | Launch · Stories | Answer only verified enrollment questions | Product FAQ + live questions | Ask an enrollment question | 0.25 | Planned | Hold until real enrollment questions exist. | |
| 11/25/2026 | Course launch | Derivative · Text post | Who the course is not designed for | Verified positioning notes | Check fit before enrolling | 0.75 | Planned | — | |
| 11/26/2026 | Internal | Course launch | Blackout · No publish | Holiday; no launch post | Supplied blackout date | None | 0 | Planned | Supplied blackout date; no publish planned. |
| 11/27/2026 | Course launch | Launch · Newsletter | Course FAQ with one closing-date reminder | Product FAQ + verified deadline | Enroll before the verified deadline | 0.75 | Planned | — | |
| 11/28/2026 | Internal | Course launch | Review · Planning | Remove unsupported claims and confirm final links | Calendar QA | None | 0.5 | Planned | — |
| 11/29/2026 | Course launch | Derivative · Reel | The onboarding mistake this month changed Maya's mind about | Month notes; write only if a real observation exists | None until the source exists | 0.8 | Idea | Needs a real observation before approval. | |
| 11/30/2026 | Internal | Course launch | Review · Retro | Record what shipped, slipped, and should be reused | November calendar states | None | 0.5 | Planned | — |
The plan uses four clusters instead of 30 unrelated prompts. Review and buffer days are visible, unsupported response-based posts stay on hold, and the final retrospective records what actually shipped. After the month, Maya would replace planned states with real outcomes rather than asking Claude to infer performance.
07 · SAME METHOD, DIFFERENT CLIENT
Use the Same Content-Planning Workflow in Codex
The editorial method does not change in Codex: provide the same input brief, use the same master prompt, require the same preview and QA report, then approve a specific Notion write—or keep the CSV fallback. There is no reason to regenerate the strategy just because the AI client changes.
To give Codex access to Notion, configure the Notion hosted MCP server in a supported local Codex environment that reads MCP configuration, then complete OAuth. OpenAI documents MCP configuration and OAuth support; Notion documents its hosted endpoint and supported tools. Access still follows the authenticated connection, so confirm the workspace and destination at the review gate.
OpenAI Developers: Model Context Protocol in CodexNotion Developers: connect an AI tool to Notion MCPNotion Developers: tools supported by Notion MCP
08 · HANDOFF DIAGNOSIS
Troubleshoot the handoff to Notion
| Symptom | Check first | Smallest safe next step |
|---|---|---|
| Claude cannot find the database | Connector/MCP authentication, selected workspace, and whether your Notion user can access the database. | Confirm the workspace and user permissions. If the connection is broader than this task should use, choose the inspected CSV path instead of assuming page-scoped MCP access. |
| Properties map to the wrong types | Exact existing names and types, especially Date, Select, Number, and Title. | Ask for a mapping-only preview and approve no row writes until every field is resolved. |
| Dates move by one day | Planning timezone, full MM/DD/YYYY values, database date values, and the CSV parser. | Export one sample with its full year, verify it in Calendar view, then import the rest. |
| CSV creates duplicates | Whether the same rows were imported before; CSV import is not a synchronization or update operation. | Add a stable source ID, inspect existing rows, and import only the missing set. |
| Accents or symbols look broken | The exported file encoding. | Export UTF-8 CSV and test a single row before the full import. |
| The calendar looks like it published | Status values and any language that confuses a Notion date with a social scheduler. | Use the shared Idea → Planned → Approved → Drafting → Ready → Scheduled → Published lifecycle, and record blockers in Notes. |
Notion Help: import data into NotionNotion Help: calendar views
09 · SOURCES AND METHOD
What is documented, and what is Kabo's method
Connector, MCP, authentication, supported-tool, import, and calendar-view statements come from the official sources below. The input brief, capacity envelope, anchor-to-derivative method, prompt, write gate, QA checklist, and Maya example are Kabo editorial material. Public creator workflows helped frame the question, but they do not prove a fixed time saving or content result and are not counted in the primary-source badge.