Imported from Canhada-Labs/ceo-orchestration (.claude/skills/domains/marketing-global/skills/twitter-engager/SKILL.md). Install upstream with npx skills add Canhada-Labs/ceo-orchestration --skill twitter-engager. Copyright stays with the author.

Twitter Engager

Cardinal Rule

Conversation precedes broadcast. The platform's organic-reach mechanism rewards replies and replies-to-replies over original tweets sent into the void. An account that publishes original content daily without allocating equivalent time to reply threads is operating against the platform's distribution logic. Engagement is the primary signal the algorithm uses to extend reach beyond the existing follower graph; every reply sent to a non-follower is a reach-extension event. Treat the reply queue as the primary publishing surface, not a secondary obligation after original posts go out.

Fail-Fast Rule

Trend participation MUST pass the topic relevance gate before any reply, quote-tweet, or thread is drafted. The gate has three conditions: (1) the trending topic intersects with the account's established subject-matter domain by direct application, not metaphor; (2) the account has prior public output on the topic that a reader could verify; (3) the tweet adds a substantive point, not a restatement of the trending narrative. If any condition fails, the correct action is to skip the trend. Forced relevance during a high-velocity trend is the most common reputational failure vector on the platform — the combination of high visibility and thin substance produces exactly the kind of negative signal that depresses future algorithmic distribution.

When to Apply

Activate this skill when the work involves any of the following:

• Designing or auditing a thread structure: hook tweet, development sequence, and call-to-action tweet.
• Evaluating whether to join a trending topic or news cycle.
• Building a reply programme targeting non-follower discovery.
• Selecting between a quote-tweet and a reply for a specific engagement objective.
• Designing a list, circle, or community architecture for an account.
• Setting or reviewing posting cadence and reply ratio targets.
• Drafting a crisis response sequence under time pressure.
• Diagnosing why an account's algorithmic reach is flat or declining.

Skip when: the task is paid Twitter/X advertising (separate SEM profile); the deliverable is a broad social calendar spanning multiple platforms (use social-media-strategist instead); or the work is Twitter Spaces audio-first programming (outside this skill's scope).

Thread Architecture

A thread is a multi-tweet unit, not a series of isolated posts. The internal structure determines whether a reader proceeds to the next tweet.

Hook Tweet

The hook tweet carries the entire load of entry and context-setting. Platform behaviour is that the hook is shown in the timeline without the thread expanded; most readers decide on the first tweet whether to tap through. The hook MUST:

• State the concrete payoff in the first line, not the last.
• Use a question or declarative assertion, not a teaser that withholds the subject.
• Avoid beginning with "I" — the algorithm has historically suppressed tweets starting with "I" in certain distribution contexts; evidence is soft but the stylistic improvement is independent of that.
• Not require domain knowledge to parse the premise — the hook's job is to earn the next tweet, not to filter for insiders.

Common failure mode: hook tweet promises a list ("5 things about X") without stating why those things matter. Replace the list promise with the insight the list proves.

Development Tweets

Development tweets (positions 2 through n-1) each carry one point. The single-point-per-tweet rule is structural, not stylistic: multi-point tweets allow readers to exit after the first point without reaching the thread's full argument. The development sequence should follow one of two valid arcs:

Numbered threads: each tweet opens with the number ("3/"). Use when the structure is a list of parallel items (tactics, rules, frameworks) and the reader benefit comes from breadth. Risk: numbering signals length; long numbered threads have higher exit rates at the midpoint.

Unnumbered narrative threads: each tweet continues a logical progression and cannot stand alone without the prior context. Use when the structure is an argument, a case study, or a sequential process. Risk: readers who enter mid-thread via a retweet lack context; mitigation is to add a reply to the final tweet pointing back to the hook.

CTA Tweet (Final Tweet)

The payoff tweet closes the loop. It MUST deliver the synthesis, not introduce new content. After the synthesis, add one engagement prompt: a question directed at the reader, a request for a specific reply, or a pointer to a related resource. The CTA tweet is the highest-leverage position in the thread for follower acquisition — readers who reach the final tweet are the most engaged subset of the distribution; a weak CTA wastes that context.

Reply Economy

Reply as Discovery Mechanism

Replies to large accounts in the same subject domain reach the audience of the account being replied to, not just the account's own followers. This is the primary organic discovery mechanism available without paid distribution. A reply that adds a new dimension to the original tweet (a counter-example, a second case, a quantification of the claim) is more likely to receive engagement than a reply that agrees or asks a question.

The mechanical practice: identify three to five accounts with audiences that overlap the target demographic; reply to their high-engagement tweets within the first 30 minutes of posting. Early replies receive disproportionate visibility in the reply sort order.

Quote-Tweet vs Reply Tradeoffs

The quote-tweet severs the reply thread, creating a separate distribution event. Use it when the comment has independent value outside the original context. Use a reply when the comment is subordinate to the original tweet and gains meaning from the parent context.

Prohibited: quote-tweeting to disagree with the content while preserving the visual amplification. This pattern delivers impressions to the quoted account and signals approval through the amplification mechanism while framing as criticism. Either reply with the disagreement or do not engage.

Reply Ratio Minimum

Maintain a minimum reply ratio of 3:1 (replies to original posts) in any rolling 7-day window. Accounts with a ratio below 1:1 appear broadcast- oriented to both the algorithm and human observers; the platform's distribution heuristics penalise low-reply accounts in timeline surfacing.

Community Building

Lists

Lists allow curating accounts for monitoring without following, and allow curated feeds to be shared publicly. Use lists for:

• Competitive monitoring: accounts in the same domain tracked without follow signal.
• Signal aggregation: high-signal accounts in a research domain grouped for rapid daily review.
• Public curation: publishing a list as a resource positions the account as a connector and attracts follower inbound from list viewers.

List maintenance discipline: review lists quarterly; remove inactive accounts (no post in 60 days); remove accounts that have drifted from the list's thematic scope.

Build-in-Public Discipline

Build-in-public content (sharing process, failures, and intermediate results) has consistently outperformed polished announcements in engagement metrics across verticals. The mechanism is: process content invites commentary before the outcome is known, generating reply threads that extend organic reach during the period when the work is in progress.

Rules for build-in-public:

1. Share a specific intermediate state, not a generic progress update ("just shipped X feature" is weaker than "the thing that almost broke us: [specific technical constraint]").
2. Do not publish proprietary customer information, material non-public commercial data, or personnel matters under the build-in-public frame.
3. Failure content outperforms success content in shares and bookmarks; resist the instinct to wait until success before publishing.

Trend Strategy

Real-Time vs Evergreen Mix

Target a 40/60 split: 40% real-time reactive content (trending topics, news commentary, live events), 60% evergreen content (frameworks, explanations, thread formats with durable value). Accounts that run above 70% real-time become unmaintainable when news cycles slow; accounts below 20% real-time miss discovery windows.

Trend-Hijack Risk Profile

Trend participation carries a risk asymmetry that most practitioners underestimate: the upside (incremental impressions on one tweet) is capped by the trend's lifespan; the downside (reputational association with a trend that later turns negative, or perception of forced relevance) can persist past the trend cycle. The topic relevance gate in Fail-Fast Rule section above is the primary control.

Additional controls:

• Timing gate: join trends within the first 2 hours or not at all. Joining a trend at peak saturation adds no discovery value and signals slow reaction time to the account's existing audience.
• Sentiment scan: before joining a breaking news or controversy trend, read 20+ replies to understand the dominant framing. A trend that appears technical may be a proxy battle for a political or personal conflict where any participation is interpreted as taking sides.
• Brand safety gate: evaluate whether the trend's dominant association conflicts with any existing brand positioning. A brand associated with reliability joining a trend about a major outage at a competitor may appear opportunistic regardless of the tweet's content.

Posting Cadence

Volume Window

3 to 10 original tweets per day is the observed range for sustainable account growth without triggering follower fatigue. Below 3, the account loses timeline presence in non-chronological feeds. Above 10, per-tweet engagement rate drops because the audience rations attention across a larger pool, and unfollows increase.

The 3 to 10 window covers original posts only. Replies, quote-tweets, and retweets do not count toward this ceiling.

Reply Volume

Reply volume is not capped at the same threshold as original posts; the platform's algorithm treats replies as engagement signals rather than publishing events. However, reply sessions should be time-bounded to avoid quality degradation. A 30-minute focused reply session twice daily (morning and late afternoon in the account's primary audience timezone) outperforms continuous unfocused reply activity.

Timing

Post original content in the 60 minutes before the account's audience peaks in activity. Platform analytics provide this data at the follower timezone level. Threads should be posted in under 15 minutes when using manual posting — long gaps between thread tweets increase the chance the algorithm surfaces a mid-thread tweet to followers who have not seen the hook, reducing thread coherence.

Crisis Response

Delete vs Correct

The default is correct, not delete. Deleting a tweet that has already received replies, quote-tweets, or screenshot circulation does not remove the content from public record — it removes the ability to provide authoritative context alongside the record that already exists. The correction tweet should be attached as a reply to the original tweet (if not deleted) or as a standalone correction citing the original tweet by timestamp and content summary (if the original was deleted for a separate valid reason such as containing personal data or a genuine safety risk).

When deleting is required: the tweet contains personal data, a material factual error that cannot be corrected without the original being visible, or content that creates a legal exposure. In these cases, delete and post a replacement immediately; a gap without replacement interpretation fills with speculation.

4-Hour Rule

A reputational incident that is not acknowledged within 4 hours generates a secondary narrative of silence or evasion. The acknowledgement does not need to include a resolution; it needs to confirm the issue is known and being addressed. Audiences on Twitter/X are explicitly aware of response time in ways that differ from other platforms; silence within the first 4 hours is consistently interpreted as negligence rather than deliberation.

The acknowledgement template: state that the situation is known, state what is being done (investigation, correction, customer contact), and give a time by which an update will be provided. Do not speculate on root cause or assign blame in the initial acknowledgement.

Transparent vs Defensive Framing

Transparent framing (acknowledging fault, describing what changed) outperforms defensive framing (explaining why the incident was not the account's fault) in sentiment recovery speed. The defensive frame may be accurate and still produce worse outcomes because it shifts the audience's interpretation from "they handled it" to "they are arguing." Transparent framing does not require admitting maximum liability; it requires avoiding language that reads as deflection.

Anti-patterns

Cross-References

• domains/marketing-global/skills/linkedin-content-creator — long-form content architecture for professional networks; use when the deliverable spans both platform strategies.
• domains/marketing-global/skills/social-media-strategist — multi- platform orchestration layer; use when Twitter/X is one channel in a broader calendar.
• domains/marketing-global/skills/reddit-community-builder — community participation disciplines; overlapping anti-patterns with Twitter reply economy (authenticity gate, lurk-before-post cadence).

ADR Anchors

• ADR-058 — bulk creative authoring constraints; skills in this domain bucket are original compositions, not translations of upstream content. The upstream source file is structural inspiration for subject coverage only; all prose, examples, frameworks, and anti-pattern tables are independently authored.