I am preparing an X campaign.

AUDIENCE:
[Describe.]

PRIMARY OBJECTIVE:
[Discovery / authority / conversation / traffic / signup / sale / affiliate action / test.]

RESOURCE / RELEASE / OFFER:
[Exact thing the campaign supports.]

CORE ANGLE:
[What I want people to understand.]

OWNED DESTINATION:
[Exact next step if relevant.]

MY VOICE:
[How I communicate.]

RELATIONSHIP RULES:
[What must remain human and should not be automated.]

AVAILABLE EVIDENCE / ASSETS:
[Projects, screenshots, articles, results, quotes, observations, examples.]

EVIDENCE SIGNAL:
[What I will use to judge the campaign.]

CONSTRAINTS:
[Time, claims, policy, disclosure, rights.]

Do not generate posts yet.

First:
1. Restate the brief.
2. Separate facts from assumptions.
3. Identify missing information.
4. Identify what should remain fixed.
5. Recommend the smallest useful X campaign scope.

Wait until the brief is coherent before building the campaign.

Review these existing X campaign assets.

Classify each as:
- OBSERVATION
- PROOF / EXAMPLE
- STANDALONE POST
- THREAD MATERIAL
- REPLY / CONVERSATION SEED
- OWNED-PATH SUPPORT
- NOT RELEVANT

Then identify the minimum missing content needed.

Do not recommend creating more content when the existing material can be adapted into the campaign.

Using the approved X campaign brief, build a 14-day operating plan.

Rules:
- Keep one campaign angle stable.
- Use a mix of observations, standalone posts, replies, proof posts and at most the threads that genuinely need sequence.
- Do not schedule posts merely to fill every day.
- Mark TEST posts and identify one variable being changed.
- State what stays fixed.
- Keep replies human.
- Identify which existing asset or observation supports each post.
- State the evidence signal.
- Include one midpoint review before planning the second half.
- Keep link-first posts in the minority unless the objective requires otherwise.

Turn this real observation into 4 X post variations.

Observation:
[Paste.]

Audience:
[Paste.]

Core angle:
[Paste.]

Keep fixed:
- meaning;
- factual claims;
- audience;
- campaign objective.

Change only:
[opening / framing / example / compression].

For each version:
1. Opening.
2. Full post.
3. Why the framing differs.
4. Evidence signal.
5. What should remain human.

Do not invent certainty, results or reactions.

Here is the idea I want to publish:
[Paste.]

Decide whether this should be:
- one standalone post;
- a longer standalone post;
- a thread.

Explain why.

If thread:
- identify the minimum number of posts needed;
- give each post one job;
- remove repetition;
- keep the strongest point early;
- avoid engagement bait.

If standalone:
- make the point complete without requiring a click.

I want to reply to this X post:
[Paste original post.]

My actual perspective:
[Paste.]

Help me prepare a reply.

Rules:
- keep my judgment intact;
- do not manufacture agreement or outrage;
- do not imitate the other person's voice;
- do not add claims I cannot support;
- do not make the reply promotional;
- keep the reply useful even if nobody visits my profile.

Give me:
1. concise version;
2. more contextual version;
3. what I should verify before posting.

Here are actual replies, quote posts, questions or objections from the X campaign:
[Paste only the relevant audience responses.]

Campaign objective:
[Paste.]

Core angle:
[Paste.]

What stayed fixed:
[Paste.]

Classify the responses into:
- repeated question;
- objection;
- confusion;
- agreement with useful detail;
- counter-example;
- qualified interest;
- unrelated noise.

Then recommend the ONE next content test with the highest learning value.

State:
1. What audience signal supports this test.
2. Whether the next piece should be a standalone post, thread, proof post, reply-led post or owned-path handoff.
3. What must stay fixed.
4. The one variable being tested.
5. The evidence signal to record.
6. What would make the result inconclusive.

Do not treat reply volume alone as proof of demand.
Do not invent audience intent.
Do not automate the relationship reply itself.

Decide whether this X campaign post has earned an owned-path handoff.

POST / IDEA:
[Paste.]

AUDIENCE:
[Paste.]

CAMPAIGN OBJECTIVE:
[Paste.]

OWNED DESTINATION:
[Article, release, signup, product, service, community or other page.]

DESTINATION PROMISE:
[What the page genuinely delivers.]

CURRENT EVIDENCE:
[Replies, bookmarks, profile visits, prior clicks or other relevant signals.]

Evaluate:
1. Can the X post stand alone without the link?
2. Does the destination add real depth?
3. Is the destination specific enough to match the post?
4. Should the link be in the post, a follow-up, profile path, or omitted?
5. What CTA is proportionate to the audience's current level of trust?
6. What evidence should I record after the handoff?

Do not force a funnel.
Do not invent urgency.
Do not recommend a link if the destination does not clearly continue the promise.

Campaign objective:
[Paste.]

What stayed fixed:
[Paste.]

Variable tested:
[Paste.]

X results:
[Impressions, bookmarks, reposts, profile visits, replies, clicks, signups, conversions.]

Qualitative feedback:
[Paste relevant replies / conversations.]

Separate:
1. OBSERVED FACTS
2. POSSIBLE EXPLANATIONS
3. UNKNOWNS

Then recommend:
KEEP
CHANGE
TEST
STOP
CONTINUE OBSERVING

Do not infer causality from one post.
Do not treat impressions as the universal success metric.
Recommend one next controlled change.

Create a reusable X campaign memory from this completed cycle.

Record:
- AUDIENCE LANGUAGE
- STRONG OBSERVATIONS
- WEAK OBSERVATIONS
- AUTHORITY THEMES
- STANDALONE VS THREAD LEARNINGS
- REPLY / CONVERSATION LEARNINGS
- LINK / OWNED-HANDOFF LEARNINGS
- CTA LEARNINGS
- TRAFFIC / CONVERSION LEARNINGS
- RELATIONSHIP WORKLOAD
- REUSABLE ASSETS
- UNRESOLVED QUESTIONS
- NEXT TEST

Label each insight:
CONFIRMED
PROMISING
UNCLEAR

Do not convert one successful post into a universal X rule.