MarketingEveryday skill
Content Calendar
Builds a realistic 4-week content calendar from your goals and audience, at a cadence you can actually keep.
What it does
Posting whenever you remember to gets you results whenever you happen to be lucky. The usual fix is a big ambitious schedule, which lasts about eleven days and then becomes guilt.
With this skill active, the assistant starts with the one goal for the next four weeks and then asks the awkward question: how many hours do you really have, and how much did you actually publish last month? The cadence comes from that answer rather than from ambition. Then it picks three or four themes that all serve the goal and lays them across four weeks, building toward your fixed dates.
Every slot comes out as something you could sit down and write: a channel, a day, a specific idea and a format, not "post about the product". Underneath sits the part most plans leave out: what to drop first when a week falls apart, when to batch the writing, and which pieces are new work versus a re-cut.
You end up with four weeks of specific, dated content ideas at a cadence your real hours support.
When to reach for it
Good for a launch, an event, a hiring push, or simply getting off the treadmill of deciding what to publish on the morning it is due. It handles teams with approval steps by scheduling the approval, not just the post.
It will not promise that four weeks of posting produces results, and if your goal genuinely needs paid advertising or an audience you do not have yet, it says so at the start rather than at the end.
What's inside
content-calendar/
├── SKILL.md
└── references/
└── examples.md
SKILL.md
---
name: content-calendar
description: Builds a four-week content calendar tied to one goal, at a cadence matched to the hours the user actually has, with a specific idea in every slot rather than a category. Use to plan posts, emails or blog content ahead of time instead of deciding what to publish on the morning it is due.
---
# Content Calendar
You build a calendar the user is still following in week three. A plan that collapses on the first busy Tuesday was not a plan, it was a wish.
## Workflow
1. **Get the five setup answers.** Goal, audience, channels, honest time budget, fixed dates.
2. **Set the cadence from capacity,** not from ambition.
3. **Pick three or four themes** that all serve the goal.
4. **Lay the four weeks out** against the fixed dates, with an arc if the goal needs one.
5. **Write a specific idea in every slot:** a topic and an angle, never a category.
6. **Add the reality clauses:** what to drop first, when to batch the writing, what gets reused.
Copy this checklist into your working notes:
```text
Content calendar checklist
[ ] One goal, stated in a sentence with a number if possible
[ ] Audience named
[ ] Channels confirmed, and the user actually posts on them
[ ] Time budget in hours per week, checked against what they published last month
[ ] Fixed dates collected
[ ] Cadence set from capacity and stated out loud
[ ] 3-4 themes chosen, each serving the goal
[ ] Every slot has a specific idea, not a category
[ ] Arc built toward the fixed dates
[ ] Drop-first order written
[ ] Batching plan written
[ ] One measure named, checked at week 2 and week 4
```
## The five setup answers
Ask for all five in one message. The fourth is the one that decides whether the calendar survives.
- **The goal for these four weeks.** One thing: signups for an event, leads, applicants for a role, awareness before a launch. If they name three goals, ask which one they would keep.
- **The audience.** Who they are talking to, and what those people already know.
- **The channels.** Only ones they will actually use. A channel added because it seems necessary is a channel that produces guilt, not content.
- **The honest time budget.** Hours per week, and then the useful follow-up: how much did you actually publish in the last month? Past behaviour is a better predictor than stated intention, and asking gently avoids building a plan for a version of the user that does not exist.
- **Fixed dates.** Launch day, event date, sale, hiring deadline, holiday closures.
If they answer three of five, build anyway and state the assumptions at the top.
## Cadence from capacity
Set the number of slots from the hours, then say the number out loud with the reasoning.
| Realistic time per week | Sustainable output |
|---|---|
| Under 1 hour | 1 post a week, one channel. Nothing else. |
| 1-2 hours | 2 posts a week, one channel, or 1 post plus 1 email a fortnight |
| 3-4 hours | 2-3 posts a week across two channels, plus a monthly longer piece |
| 5+ hours, or a team | 3-4 posts a week, a weekly email, and one substantial piece a month |
Two things worth saying plainly: an email newsletter costs roughly three times what a social post costs, and video costs more than both. And two solid posts a week beat seven rushed ones, because the rushed ones stop after eleven days.
When the user asks for more than their hours support, say what would have to be true (batching, reuse, or someone else writing) rather than agreeing and building a plan that fails.
## Themes that carry a goal
Pick three or four. Every slot belongs to one of them, and every theme has a job.
- **Teach.** Something useful the reader can act on without buying anything. Builds the reason to follow.
- **Prove.** Evidence that the thing works: results, a case, a before and after, a customer's words.
- **Show.** How the work actually happens: the process, the people, the decisions. Builds trust more cheaply than either of the above.
- **Invite.** The ask: register, apply, reply, buy. This is the theme people under-use, then wonder why nothing converts.
A rough mix that works for most four-week plans: half teach, a quarter prove, a quarter split between show and invite. Weight it toward invite in the week before a deadline. This is a starting point, not a law.
## The four-week arc
The arc depends on the goal. Say which one you used, in a line.
| Goal | Week 1 | Week 2 | Week 3 | Week 4 |
|---|---|---|---|---|
| Event or launch on a date | Plant the problem | Teach around it | Proof and open registration | Final push and last call |
| Steady growth, no deadline | No arc. Consistent themes, one recurring format that becomes recognisable | | | |
| Hiring | What the team is working on | How the team works | The people already there | The role and how to apply |
| Awareness before a bigger campaign | Broad and useful | Narrow to your angle | Show the work | Introduce what is coming |
Anchor the arc on the fixed dates first, then fill around them. A launch post that lands two days after launch is worse than one that lands two days before.
## Write slots, not categories
A slot the user cannot sit down and write is not a plan, it is homework.
| Category | Slot |
|---|---|
| Post about the product | The three questions every customer asks in week one, and our answers |
| Share a case study | What the Beckett job cost us to learn: the quote we got wrong and why |
| Thought leadership piece | Why we stopped offering the cheapest option, and what happened to enquiries |
| Behind the scenes | A photo of the whiteboard from Monday's argument about the pricing page |
| Newsletter | Three things we changed after reading our own support tickets |
Each slot needs a channel, a day, the idea in one line, and the format. If you cannot write the idea in one line, the idea is not there yet.
## Reuse, because it is the only way this survives
One real idea should produce three publications, and this is not cheating, because almost nobody sees everything you post.
- A long post becomes an email, a short post, and a comment on someone else's thread.
- A customer question becomes a post, an FAQ entry and a slide.
- A month of posts becomes one substantial piece at the end of it.
Mark reuse explicitly in the calendar so the user sees which slots are new writing and which are re-cuts. A calendar of twelve original pieces is a different commitment to one with five originals and seven re-cuts.
## Batching and the drop-first order
**Batching:** one writing block produces a week or two of content far more reliably than twelve separate moments of inspiration. Suggest a specific block ("Thursday morning, two hours, write the following week") rather than the general advice.
**Drop-first order:** list, in order, what goes when a week falls apart. Usually: the reused piece first, then the show slot, then a teach slot. Never the invite slot in the final week, and never the piece tied to a fixed date. Writing this down converts a collapsed week from a failure into a plan working as designed.
**Slack:** leave one slot a week empty on purpose, or plan three weeks of content across four. Every calendar meets a week it did not expect.
## Measuring, honestly
Pick one number that matches the goal (registrations, replies, applicants, subscribers) and check it at the end of week two and week four.
Be straight about what four weeks can tell you: it will show you which posts got attention and whether the goal number moved, and it will not prove why. Treat it as direction, not proof. The most useful output is usually the list of which two ideas landed, because that shapes the next four weeks better than any engagement average.
## Example
**Setup:** solo consultant, goal is 50 registrations for a webinar on 28 May, audience is operations managers in mid-sized manufacturers, channels are LinkedIn and a 900-person email list, two hours a week, and they published four times in the last month.
**Cadence:** two LinkedIn posts a week plus one email a fortnight. That is the honest ceiling for two hours, and it matches what they are already managing.
**Themes:** teach, prove, invite. No "show" theme, because a solo consultant has less behind-the-scenes material and it costs time they do not have.
**Week 1: plant the problem**
- [LinkedIn, Tue] The most expensive line on a manufacturing shift report, and why nobody reads it. (teach)
- [LinkedIn, Thu] What three plant managers told me when I asked why the report exists. (teach)
**Week 2: teach around it, open registration**
- [Email, Tue] The shift-report problem, written out properly, ending with the webinar date and link. (teach + invite)
- [LinkedIn, Wed] Registration is open, with the one specific thing attendees will leave with. (invite)
- [LinkedIn, Fri] Re-cut of the email's middle section as a short post. (reuse)
**Week 3: proof**
- [LinkedIn, Tue] The Halden plant: what changed in the first six weeks after they rebuilt the report. (prove)
- [LinkedIn, Thu] The part of that project that did not work, and what I would do differently. (prove)
**Week 4: final push**
- [Email, Mon] Last call, with the agenda and the time zone stated. (invite)
- [LinkedIn, Tue] Three days left, and the one question I get asked most about this. (invite)
- [LinkedIn, Thu, morning of the webinar] Doors close at 2pm, here is what we are covering. (invite)
**Drop first if the week collapses:** the Week 1 Thursday post, then the Week 2 Friday re-cut. Never drop the Week 4 email or the day-of post.
**Batch:** write Weeks 1 and 2 in one two-hour block before Week 1 starts. Week 3's proof posts need the Halden numbers, so gather those first.
**Measure:** registrations, checked after Week 2 and again on the 27th. If the number is under 20 after Week 2, the fix is a direct ask to the email list, not more posts.
## Rules
- Match the cadence to the hours the user actually has, and say the number out loud.
- Every slot serves the goal. If it does not, cut it rather than finding a justification.
- Specific ideas only. "Post about the product" is not a slot.
- No invented trends, fake awareness days or seasonal hooks that do not exist.
- Only formats the user can actually produce. Do not plan video for someone who has never made one.
- Leave slack. A calendar with no gap does not survive contact with a real week.
- Say what to drop first, in order, before the busy week arrives.
- Never promise results from a four-week plan. Content compounds slowly, and saying so builds more trust than a forecast.
## Edge cases
- **No real goal, just "grow".** Push once for something countable: subscribers, enquiries, applicants. If there genuinely is not one, build for consistency and one recurring format, and say plainly that this is a foundation plan rather than a campaign.
- **They want to post daily.** Ask what they published last month. If the answer is four things, a daily plan will produce ten days of content and a month of guilt. Offer the sustainable cadence and a route to daily later.
- **A team with approvals.** Build the approval time into the calendar as its own slot. A post needing two sign-offs must be written a week before it publishes, and the calendar should show that, not pretend it publishes the day it is written.
- **A regulated industry.** Finance, health, legal and similar: compliance review is a real step with real timing. Put it in the calendar, keep claims out of the ideas, and say that the wording of anything making a claim needs review before it goes out.
- **No audience yet.** A calendar will not create one on its own. Be honest: the first four weeks are about having something to point at. Weight it toward one format, and add a distribution slot (commenting, sharing in communities, direct outreach), because publishing into silence is the most common way people conclude that content does not work.
- **A single-person business in a busy season.** Plan three weeks of content across four and say why. A realistic plan they finish beats an ambitious one they abandon in week two.
- **The goal needs paid advertising, not content.** Say it. Fifty registrations in four weeks from a 900-person list is plausible; five thousand is not, and no calendar will fix that.
- **A channel you do not know well.** Say so, keep the ideas and the cadence, and let the user adjust the format conventions. Confident advice about a platform's specifics that turns out to be wrong costs more than the admission.
## Files
- `references/examples.md` - three complete four-week calendars: a solo consultant driving event registrations, a local business hiring, and a product team running a launch through an approval process. Each one shows the cadence reasoning, every slot, and the drop-first order.
Install it
Claude Code (this project)
Unzip the download into .claude/skills/. The folder is already named right.
Claude Code (every project)
Unzip into ~/.claude/skills/ to use it everywhere.
claude.ai
Upload the zip under Settings, then Capabilities, then Skills (paid plans).
How to call it
Just describe the task. The description line inside the skill does the matching. Or call it directly:
/content-calendar
make me a content calendarplan four weeks of postswhat should we post before our launchbuild a posting schedule I can actually keepI keep forgetting to post, help me plan ahead
Works beyond Claude
Skills use the open Agent Skills format. The exact same folder works in OpenAI Codex, Gemini CLI, Cursor, GitHub Copilot and other tools that read the format.
Using a plain chat app? Paste SKILL.md into ChatGPT or Gemini as instructions. You lose auto-triggering, but the method still works.
agentskills.io