

A specialist agent is only as good as the brief it reads. The brief is the contract between the human strategist and the drafter. After a year of iterations, this is the six-field brief template that runs every weekly content cycle inside Arthea. Six fields. Copy it into your own pipeline and you will get most of the lift the schema is designed to deliver, regardless of which model or platform you draft on.
Most teams that ship weekly content fight the same problem. Drafts are uneven. Voice drifts. Reviewers spend more time fixing tone than fixing argument. The cause is almost always upstream: the brief was thin, or the brief was a paragraph of vibes, or the brief was a Slack message that the drafter half-read. The six-field brief template is the structural fix. It does not make a bad strategist a good one. It makes a competent strategist consistently shippable, week after week, on the same brand voice contract.
The six fields in the weekly brief template
Six fields, in this order: Pillar, Voice, Hook angle, Key points, CTA, References. Each field has a job. Each is a hard requirement before a draft starts. The schema is the boring part on purpose; the strategy lives in the bullets.
Pillar
Which content theme this post belongs to. P1 product, P2 industry, P3 builder, P4 ops, P5 resource, P6 community is one workable schema; brand-specific pillars are equally valid as long as the set is fixed and named. Pillar drives image template, destination URL, and how aggressively the editor scores the draft against brand-direction guardrails. Pillar is routing metadata. It never appears in the copy itself.
The reason pillar comes first is that downstream automation reads it. Image generation picks the right template, scheduling picks the right slot in the weekly cadence, the analytics layer attributes performance to the pillar so a strategist can see at the end of the quarter which pillars carried the engagement and which underperformed. None of that downstream value lands without the pillar field being filled correctly at brief time.
Voice
One of: operator, contrarian, builder, founder, teacher. Each voice has a different register and rhetorical posture. Operator is direct, low-adjective, mechanism-first. Contrarian leads with the false-consensus claim and dismantles it. Builder is process-oriented, sequenced, technical. Founder is reflective, biographical, lower-frequency. Teacher unpacks one concept clearly with named parts.
Pin the voice in the brief so the same brand reads the same way regardless of which agent or human drafted the post that week. Voice consistency at the schema layer is what lets a brand build a recognisable presence across thousands of posts; voice drift at the brief layer compounds into a brand that sounds like five different people and reads as inauthentic to the audience.
Hook angle
One sentence naming the lens. The angle is the strategic frame; the headline emerges from drafting against it. "Why hourly billing breaks under AI" is an angle. "10 reasons hourly billing is dead" is a headline. The angle stays in the brief; the headline emerges in the draft. Confusing the two is the most common failure mode for new strategists, and the schema enforces the distinction.
A good angle does three things at once. It names a tension the audience already feels but has not articulated. It commits to a position, not a survey. And it can be argued against. If the angle could be flipped and still feel true, it is not yet sharp. The strategist iteration on the brief is mostly iteration on the angle.
Key points
Two to four bullets, each with a number, a name, or a mechanism. The drafter expands them into prose. The strategist job is to make the bullets specific. Generic bullets produce generic copy. This is the highest-leverage field in the schema. A brief with weak key points cannot be saved by a strong drafter; a brief with strong key points produces shippable output even from a mediocre one.
The bullets are not the post outline. They are the substance the drafter must include. Order, transitions, framing, and rhetorical structure happen in the draft. The bullets are the load-bearing claims, each one with a hook the drafter can grip on: a number, a named mechanism, a specific example. "Engagement compounds" is not a key point. "Welcome conversion drops measurably for every additional hour after signup" is.
CTA
The next-step action you want a reader to take. Read more, book a call, download something, follow. This determines the closing line and the URL. CTA is one field because varying it across posts dilutes the brand call to action; pinning it in the brief forces the strategist to be explicit about what this specific post is for inside the larger content arc.
References
Optional but recommended. URLs to whatever informed the angle: an article you read, a post you are responding to, a paper, an internal Atlas Library entry, a transcript. The drafter cites or paraphrases. References also help dedup so two posts from the same source do not collide in the calendar. The reference field is also the audit trail for the strategist; six months in, the references explain why the post existed when nobody remembers the original moment that prompted it.
What is not in the weekly brief template
Word count. Tone modifiers ("punchy", "punchier", "make it pop"). Audience persona paragraphs. Anything more verbose than the post you are trying to produce. If the brief takes more than 90 seconds to write, the shape is wrong and you are inventing a problem the schema does not have.
Word count belongs to the platform contract, not the brief. The drafter knows the X length, the LinkedIn length, the Threads length. Tone modifiers belong to the voice gate, not the brief; "punchy" is what an operator says when the voice field was left vague, and the right fix is to tighten the voice contract, not to layer on subjective adjectives at brief time. Persona paragraphs belong to the brand-direction document the agent reads at setup; they are global context, not per-brief input.
Why six fields is the equilibrium
Fewer than six and the agent has nothing to ground in. More than six and the brief stops being writable in the available time. Six is the equilibrium for repeated weekly content with a consistent voice contract. The schema has been tested against shorter and longer variants; six survives the cadence test and the quality test simultaneously, which is why it is the published shape.
Every layer of the system reads the same six fields. The drafter writes against them, the voice scanner audits compliance with them, and the publisher refuses outputs that drift from them. Consistency at the schema layer is what lets a specialist agent ship cadence-rate output without per-draft supervision. The schema is the contract that makes parity-quality output cheap.
Worked example of the six-field brief template
Illustrative, not a real brief. Pillar: P3 builder. Voice: operator. Hook angle: most retention dashboards are healthy on metrics that lag the regression by ninety days. Key points: 1) attributed flow revenue per active subscriber is the only metric that catches deliverability slips early, 2) Apple Mail Privacy makes open rate diagnostic-grade only in aggregate, 3) cohort margin moves before headline conversion when discount stacking starts conditioning the list, 4) segment freshness is invisible on every standard dashboard. CTA: read the full audit deep dive. References: the Klaviyo deep-dive Architecture Notes article, two internal audit notes. This brief takes about sixty seconds to write and produces a draft that is shippable with one editing pass.
Runbook: adopting the weekly brief template in your pipeline
1. Copy the schema into whatever doc tool the strategist uses today. Do not migrate the team to a new tool just for this. 2. Test on one specialist channel first. LinkedIn and X are the easiest to baseline because cadence is high and feedback loops are fast. 3. Run it for two weeks before changing anything. Two weeks is the minimum window to separate brief-quality signal from drafter-quality noise. 4. Iterate on the bullets. Keep the schema fixed. The schema is the boring part on purpose; the strategy lives in how specific the key points get. 5. Audit one drafted post a week against the brief that produced it. Note where the draft drifted from the brief. Drift always traces back to a weak field upstream. 6. Once the cadence is stable, expand to the next specialist channel using the same schema. Do not fork the schema per channel; the schema is what keeps the brand consistent across channels.
When the six-field brief template is wrong
The schema is calibrated for repeated weekly content with a consistent brand voice. It is wrong for one-off long-form pieces (a cornerstone article, a research report, a launch keynote), which need a richer brief because the cadence does not amortise the schema cost across enough posts. It is also wrong for brands that are still discovering their voice; the voice field assumes the contract is already written, and a brand pre-voice should ship five or ten one-off posts in different registers first to learn what works before committing the schema.
The other anti-pattern is treating the schema as documentation rather than input. A brief that gets filled out for compliance and then ignored by the drafter is worse than no brief at all, because it adds process cost without producing structural lift. The schema only earns its keep when the drafter and the voice gate both read from it.
What success looks like
On a brand that adopts the six-field brief template and runs it for a quarter, the visible signals are: brief writing time stable around sixty to ninety seconds, drafter editing rounds dropping to one per post, voice gate flag rate dropping into the low single digits, and reviewer time shifting from tone fixes to substance review. The qualitative signal is the strategist who, three months in, can write a brief on a five-minute walk and know it will produce a shippable draft.
On the AI Lab side of the engagement, the published outcome bands are 20 percent retention lift and 60 percent operator time saved on the work the schema enables. The brief is what makes that lift possible because every downstream layer of the system depends on it being consistent.
FAQ
Why six fields and not five or seven? Fewer than six and the drafter has nothing to ground in. More than six and the brief stops being writable in the available time. Six is the equilibrium for repeated weekly content with a consistent voice contract.
Can the same six-field brief template work across X, LinkedIn, Threads, and Instagram? Yes, that is the design. The schema is platform-agnostic; only the platform contract on length and format changes downstream. Forking the schema per channel is the anti-pattern.
How long should writing one brief take? Around sixty to ninety seconds once the voice contract is set. If the brief takes longer than that, the strategist is inventing a problem the schema does not have, usually by trying to write the post inside the brief.
What is the highest-leverage field? Key points. Generic bullets produce generic copy. Specific bullets, each with a number, a name, or a mechanism, produce shippable output even from a mediocre drafter.
Where does the brief sit in the AI Lab content stack? At the top. The brief is the contract every downstream layer reads from: drafter, voice gate, publisher, analytics. Engagement detail is on /ai-lab.
Read more
- https://www.arthea.ai/ai-lab - https://www.arthea.ai/article/instrumenting-ai-content - https://www.arthea.ai/article/five-phase-webflow-cro-architecture
If you want a 30-minute review on adopting the six-field brief template inside your weekly content cycle, the calendar is at arthea.ai/book.
Related: An attribution model for content that compounds over months
Related
- The four signals every AI content program should log from day one
- The five-step brief-to-ship process for AI content cadence
- The four numbers we check every morning
- The weekly shipping log is the most underrated marketing artifact we publish
- The one-prompt drift classifier we run on every AI draft
- The three metrics that predict whether your content works

Architecture Notes
Occasional insights on infrastructure, conversion systems, retention architecture, and AI deployment, shared when they’re worth reading.











