Agent skill

Launch Window Mechanics

by swan-gtm in swan-gtm/gtm-skills

A skill your agent uses when a post is about to go live, or just went live, and the question is what happens after publish — a creator's sponsored post, a founder announcement, a launch thread, a…

MITAuto-check passedProduct & Project Management

Install Launch Window Mechanics

skills CLI
$ npx skills add swan-gtm/gtm-skills --skill launch-window-mechanics -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install swan-gtm/gtm-skills launch-window-mechanics --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/swan-gtm/gtm-skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/danni-chen/launch-window-mechanics .claude/skills/launch-window-mechanics && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
launch-window-mechanics
GitHub stars
172
Token cost
~2.8k tokens
SKILL.md length
1,502 words
Files
6 (incl. references)
Skills in repo
32
Repo updated
First seen
Licence
MIT

At a glance

A skill your agent uses when a post is about to go live, or just went live, and the question is what happens after publish — a creator's sponsored post, a founder announcement, a launch thread, a…

  • A post is about to go live
  • SKILL.md covers Before the post exists, Interaction leverage, The sequence and Amplification: only behind…, plus 4 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
  • The question is what happens after publish — a creators sponsored post

What it does

Launch Window Mechanics is an agent skill from swan-gtm/gtm-skills. Use this skill when a post is about to go live, or just went live, and the question is what happens after publish — a creator's sponsored post, a founder announcement, a launch thread, a milestone update, a customer story. "The post is scheduled, now what", "why did our sponsored post die", "should we put spend behind this one", "how do we get a post moving in the first hour", "the creator delivered and nothing happened", "we got impressions but no pipeline", "is this post working". Produces the pre-publish…

Its SKILL.md is about 2.8k tokens, which your agent loads only when the skill is triggered. The skill folder holds 6 other files, including reference files (for example `references/amplification.md`, `references/consumer-ai.md` and `references/creator-scorecard.md`).

It sits in Product & Project Management, covering Storytelling and Project management. The repository describes itself as: Open, production-grade GTM skills for AI agents. The licence is MIT.

When your agent uses it

  • A post is about to go live
  • The question is what happens after publish — a creators sponsored post
  • A founder announcement
  • A launch thread

Example prompts

  • “The post is scheduled, now what”
  • “why did our sponsored post die”
  • “should we put spend behind this one”
  • “/launch-window-mechanics”

What it can do on your machine

Read from SKILL.md and the folder at commit 67abd04. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Launch Window Mechanics loads about 2.8k tokens when it runs, and up to ~8.1k if it reads all its reference files. Until then it costs about 180 tokens; SKILL.md has 1,502 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~180
When it runs · the whole SKILL.md, loaded when a task matches
~2.8k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~8.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from swan-gtm/gtm-skills at commit 67abd04, republished under its MIT licence (© swan-gtm). 1,502 words, ~2,774 tokens.

Download SKILL.mdSave it as .claude/skills/launch-window-mechanics/SKILL.md (or your agent's skills folder). This skill also uses 5 other files; get the full folder from GitHub.
name
launch-window-mechanics
description
Use this skill when a post is about to go live, or just went live, and the question is what happens after publish — a creator's sponsored post, a founder announcement, a launch thread, a milestone update, a customer story. "The post is scheduled, now what", "why did our sponsored post die", "should we put spend behind this one", "how do we get a post moving in the first hour", "the creator delivered and nothing happened", "we got impressions but no pipeline", "is this post working". Produces the pre-publish setup, the minute-by-minute sequence from T−30 to T+120, the paid amplification trigger and its stop conditions, the debrief, and the read on whether a post is alive or already dead.
title
Launch window mechanics
category
Influencers
tags
Marketing, Demand Gen

Launch Window Mechanics

Feeds do not distribute a post. They test it. A small slice of the audience sees it first, the system scores the reaction over the following half hour, and distribution either widens or stops. Almost everything written about creator marketing ends at publish. The reach was decided after that.

The job in this window is narrow: produce enough real, fast, high-quality reaction inside the test to clear the threshold, then decide whether the result deserves paid spend behind it.

Produces: a pre-publish setup, a timed sequence from thirty minutes before to two hours after, a go/no-go call on amplification with stop conditions, and a debrief that feeds the next cycle.

Before the post exists

The window cannot be worked if the post was not built to be worked. Three things are decided at draft time, not at publish time.

The post must create an obligation to respond. A neutral summary earns scrolling. A stated position, an open question, or a claim someone would want to correct earns replies, and replies are what the ranker weighs. This is not a call for provocation. It is the difference between "here is what we shipped" and "most teams are still doing this the slow way, here is the number that changed our mind."

Rights and attribution are contractual, not operational. The creator agreement must grant paid usage rights on the post itself, with a stated window. Without that clause, the post you spend six hours validating cannot be amplified, and renegotiating mid-window costs both time and leverage. This is the most common preventable failure in the sequence. Attribution goes in at the same time: a distinct tracking parameter or code per creator, and where links are not permitted, a read on branded-search lift instead.

The traffic needs somewhere to land. Reach is the entry, retention is the business. A spike arriving at a page with no next step is a spike that churns. Decide the destination before creating the demand: a signup path, a template, a resource, a community.

Choosing who to run this with: references/creator-scorecard.md.

Interaction leverage

Reactions are not worth the same, and the gaps between them are much larger than most teams assume. Ordered by leverage, strongest first:

SignalWhy it scoresWhat it means operationally
Author replying to each commenterEvidence of a live conversation rather than a broadcastThe highest-value action available, by a wide margin. Sit in the comments and answer individually
Substantive replies from othersCosts real effort, therefore hard to fakeBuild the tension or open question into the post so that replying is the natural response
Reshares carrying added commentaryRedistributes into a graph the post could not otherwise reachLine up two or three people who will add a genuine opinion, not agreement
Saves and bookmarksThe silent endorsement, and a strong quality proxyLists, data, and frameworks earn these; opinion posts rarely do
Profile clicksThe post made someone curious about the authorA byproduct of a clear point of view, not something to engineer
LikesThe weakest signal on offerLikes with no replies is a recognised low-quality pattern
Hides, mutes, reportsStrongly negative, and suppressiveExactly what purchased engagement produces

Two structural constraints apply nearly everywhere. Outbound links in the post body suppress distribution, so the link goes in the first comment. Verified accounts receive materially wider reach than unverified ones, which makes verification a cost of entry rather than an upgrade.

Weightings, thresholds, and per-surface differences: references/platform-signals.md.

The sequence

WhenAction
T−30Publish into the hours the target audience actually reads, not the hours convenient for the team. Draft the first comment carrying the link. Write three to five seed questions that would each pull a second person into the thread. Tell the people who genuinely care that it is about to go up
T+0–15Post the link comment immediately. The author answers every comment as it lands, individually, in substance. Target roughly ten substantive reactions inside fifteen minutes. This block clears the first threshold or it does not
T+15–60The author keeps replying and extends threads rather than closing them with thanks. Reshares with commentary land here. Post a second question to restart the conversation once the first wave settles
T+60–120Keep replying. Check the amplification trigger. Cross-post to a second surface for a second pulse rather than a duplicate
T+6–8hThe amplification decision resolves. Trigger met and sentiment positive, deploy. Otherwise let it run
T+2dPublish the process teardown: how it was actually made, what it cost, what broke. Reliably produces a second wave, and it is planned in the original brief rather than improvised
T+3–7dPublish the method or template and the getting-started path, so viewers become participants

Never edit the original after publish. Never post and leave.

Amplification: only behind proven winners

Paid spend does not rescue a weak post. It widens one that organic distribution already validated. The post itself is boosted so it stays native, rather than rebuilt as an ad, which discards the social proof that made it work in the first place.

Trigger. Engagement reaches roughly 1.5 to 2x the account's own trailing baseline, within six to eight hours of publish, with sentiment clearly positive. The baseline is that account's own recent performance, never a cross-industry benchmark, because a strong post on a small account and a weak post on a large one produce identical absolute numbers.

Stop. Cut on any one of these, not on all: cost per action running thirty percent over target for two consecutive days; frequency climbing while click-through declines; organic engagement on the original falling below its own baseline; a cluster of brand-safety-negative comments.

A trigger stated in numbers is what separates a decision from a habit. Without one, spend goes behind whatever the team happens to like.

Four-layer model, budget caps, test design, and a worked example: references/amplification.md.

Show full SKILL.md (526 more words)Show less

Pre-publish check

  • The post carries a clear interaction trigger, an open question or a stated position, not a neutral summary
  • Zero outbound links in the body; link drafted and ready for the first comment
  • No more than two hashtags
  • The publishing account is verified
  • Publish time matches when this audience actually reads
  • Three to five seed questions written; the people who care have been told it is coming
  • Paid usage rights confirmed in the creator agreement, with a stated window
  • Attribution wired: distinct parameter or code per creator, branded-search read where links are disallowed
  • The destination exists: signup path, template, or resource
  • The author has the first hour blocked out and is not in meetings

What good looks like

  • Reshare speed is the call, not like count. When reshares start outrunning likes, the post has crossed from something people consume to something people forward. That crossing is the earliest reliable read, and it appears well before the totals look impressive.
  • Read who reacted before how many. Twenty substantive replies from an adjacent community beat two hundred likes from anywhere, and engagement from the wrong audience scores against the post rather than for it. If nobody from the target buying group is in the first wave, more volume will not repair it.
  • Likes with no replies is a failure state, not a modest success. It is the exact pattern a ranker treats as low quality, and it is what purchased engagement produces.
  • Sentiment is a gate, not a footnote. Positive above roughly eighty percent, negative below five. Attention carrying a negative split costs more than it returns, and amplifying it buys a larger argument.
  • Reach is the entry, retention is the business. A spike that lands nowhere is a spike that churns.
  • The three failures, in order of frequency: publishing and walking away; buying engagement; and putting spend behind a post that never cleared its own organic baseline.
  • Scoring differs by motion. High-volume consumer runs on hit rate across many posts. A small-roster motion cannot, because the denominator is too small for the ratio to mean anything and the payoff is lagged pipeline rather than views. See references/consumer-ai.md.
  • Review cadence, metrics, and the trap list: references/debrief.md.

Rules

  • MUST have the author personally answer comments through the first hour. Nothing else in this skill compensates for skipping it.
  • MUST use only real people on their own accounts, reacting because they have something to say. Coordinated pods, reciprocal-engagement groups, and purchased engagement are excluded, on quality grounds as much as ethical ones: they generate the exact low-quality pattern the ranker penalises.
  • MUST confirm paid usage rights before publish, or the validated winner cannot be amplified.
  • MUST set the amplification trigger and stop conditions in advance, in numbers.
  • MUST disclose commercial relationships according to the applicable rules. Undisclosed sponsorship is a legal exposure and an audience one.
  • NEVER put the link in the post body.
  • NEVER boost a post that has not cleared its own organic baseline.
  • NEVER edit the original after publish.
  • NEVER give a creator material that performs far beyond what an ordinary user can reproduce. The gap becomes a credibility problem that outlasts the campaign.

Example prompts

The post goes live in an hour. Walk me through the setup.
Six hours in: engagement is 1.4x our average, sentiment is mixed. Boost or hold?
Our creator post got 900 likes and 4 comments. Is it working?
Score these three creators for a launch we need to work hard in the first hour.

© swan-gtm, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 5 other files (references) in skills/danni-chen/launch-window-mechanics of swan-gtm/gtm-skills.

  • SKILL.md
  • references/amplification.md
  • references/consumer-ai.md
  • references/creator-scorecard.md
  • references/debrief.md
  • references/platform-signals.md

Open the folder on GitHubat commit 67abd04

Compare with similar skills

Launch Window Mechanics next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Launch Window Mechanics compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Launch Window Mechanics this skillswan-gtm/gtm-skills172—~2.8kAutomated safety check: PassMIT
Oral Paper SkillAdkid-Zephyr/oral-paper-skill357—~1.9kAutomated safety check: PassNone
Prd V10 Case Study Buildermattgierhart/PRD-driven-context-engineering180—~2.9kAutomated safety check: PassMIT
Operating System Visualizationmingchen666/Reviva244—~1.2kAutomated safety check: PassNone
CCPM Project Managementautomazeio/ccpm8.4k—~1.1kAutomated safety check: PassMIT
Uvastral-sh/claude-code-plugins3132 repos~980Automated safety check: PassApache-2.0

Similar skills

  • Oral Paper Skill

    Adkid-Zephyr/oral-paper-skill

    Help authors learn from exemplary ICLR, ICML, and NeurIPS papers through source-linked manuscript comparisons, concrete writing and experiment suggestions, and guided reflection.

    357 GitHub stars~1.9k tokensUpdated 23 days ago
    Product & Project ManagementAuto-check passed
  • Prd V10 Case Study Builder

    mattgierhart/PRD-driven-context-engineering

    Build customer case studies as marketing and chasm-crossing reference assets during PRD v1.0 Market Adoption.

    180 GitHub stars~2.9k tokensUpdated 1 mo ago
    Product & Project ManagementAuto-check passed
  • A skill your agent uses whenever the user wants visual learning for operating systems, process states, CPU scheduling, Gantt charts, synchronization, semaphore/PV problems, deadlock, Banker's…

    244 GitHub stars~1.2k tokensUpdated yesterday
    Product & Project ManagementAuto-check passed
  • Runs a spec-driven workflow from PRD to epic to GitHub issues to parallel agents, with status, standup and blocked-work reports from bundled scripts.

    8.4k GitHub stars~1.1k tokensUpdated 6 mo ago
    Product & Project ManagementAuto-check passed
  • Uv

    astral-sh/claude-code-plugins

    Official

    Guide for using uv, the Python package and project manager. An agent skill from astral-sh/claude-code-plugins.

    313 GitHub starsUsed in 2 repos~980 tokens
    Product & Project ManagementAuto-check passed
  • Project Management

    kunchenguid/firstmate

    Agent-only procedure for Firstmate project management. An agent skill from kunchenguid/firstmate.

    7.8k GitHub stars~2.1k tokensUpdated today
    Product & Project ManagementAuto-check passed

More from swan-gtm/gtm-skills

All 32 skills in this repo
  • Revops Revenue Planning

    swan-gtm/gtm-skills

    Annual and quarterly revenue plan construction, top-down vs bottoms-up reconciliation, plan versioning, stretch goal handling, and FP&A-RevOps collaboration for B2B revenue teams.

    172 GitHub stars~7.2k tokensUpdated 3 days ago
    Auto-check passed
  • AI Personalization Prompts

    swan-gtm/gtm-skills

    A skill your agent uses when setting up AI-powered personalization, building Clay or lemlist workflows, or automating prospect research — 6 AI personalization prompts (lemlist style) plus 2 email…

    172 GitHub stars~653 tokensUpdated 3 days ago
    Auto-check passed
  • Audience Icp Filter

    swan-gtm/gtm-skills

    A skill your agent uses when a list of people already exists and someone needs to know who on it is worth contacting — event or webinar attendees, registrants, a prospecting export, a CRM segment, a…

    172 GitHub stars~1.9k tokensUpdated 3 days ago
    Auto-check passed
  • Brand Mention Monitor

    swan-gtm/gtm-skills

    A skill your agent uses when you need to know what people are saying about a brand across the web and social — "monitor brand mentions", "what are people saying about [brand] this week", "run a…

    172 GitHub stars~2.1k tokensUpdated 3 days ago
    Auto-check passed
  • Bridge Before Cold

    swan-gtm/gtm-skills

    Use this skill before staging a prospect and before drafting any first touch, when a segment has gone silent, and when deciding whether an account is genuinely cold.

    172 GitHub stars~1.4k tokensUpdated 3 days ago
    Auto-check passed
  • Champion Move Detection

    swan-gtm/gtm-skills

    Use this skill on a monthly cadence to detect champions and heavy users of your paying customers who changed jobs, verify the move against live LinkedIn data, score the new company, and surface…

    172 GitHub stars~5k tokensUpdated 3 days ago
    Auto-check passed

Questions about Launch Window Mechanics

What does Launch Window Mechanics do?

A skill your agent uses when a post is about to go live, or just went live, and the question is what happens after publish — a creator's sponsored post, a founder announcement, a launch thread, a…. Launch Window Mechanics is an agent skill from swan-gtm/gtm-skills. Use this skill when a post is about to go live, or just went live, and the question is what happens after publish — a creator's sponsored post, a founder announcement, a launch thread, a milestone update, a customer story.

When should I use Launch Window Mechanics?

Launch Window Mechanics fits situations like: A post is about to go live; the question is what happens after publish — a creators sponsored post; A founder announcement; A launch thread.

How do I install Launch Window Mechanics in Claude Code?

Run `npx skills add swan-gtm/gtm-skills --skill launch-window-mechanics -a claude-code`. Or copy the skill folder (skills/danni-chen/launch-window-mechanics in swan-gtm/gtm-skills) into .claude/skills/launch-window-mechanics in your project. Claude Code loads it when a task matches its description.

How do I install Launch Window Mechanics in Codex?

Run `npx skills add swan-gtm/gtm-skills --skill launch-window-mechanics -a codex`. Or copy the skill folder (skills/danni-chen/launch-window-mechanics in swan-gtm/gtm-skills) into .agents/skills/launch-window-mechanics in your project. Codex loads it when a task matches its description.

Can I use Launch Window Mechanics in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add swan-gtm/gtm-skills --skill launch-window-mechanics -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/launch-window-mechanics, .gemini/skills/launch-window-mechanics, .github/skills/launch-window-mechanics and .opencode/skills/launch-window-mechanics in your project.

What does Launch Window Mechanics need to run?

SKILL.md names no scripts, command-line tools or credentials: Launch Window Mechanics is instructions for the agent only.

Does Launch Window Mechanics access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Launch Window Mechanics safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Launch Window Mechanics use?

Launch Window Mechanics is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Launch Window Mechanics use?

About 2.8k tokens (SKILL.md is roughly 11k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 5.4k tokens, read only when the agent opens those files.

What are the alternatives to Launch Window Mechanics?

Skills that share tags, products or a category with Launch Window Mechanics: Oral Paper Skill (Adkid-Zephyr/oral-paper-skill, 357 stars), Prd V10 Case Study Builder (mattgierhart/PRD-driven-context-engineering, 180 stars), Operating System Visualization (mingchen666/Reviva, 244 stars) and CCPM Project Management (automazeio/ccpm, 8.4k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Launch Window Mechanics?

swan-gtm (a GitHub organization) maintains it in swan-gtm/gtm-skills, which has 172 GitHub stars. The repository holds 32 skills in this directory. The repository was last updated on October 8, 2026.

Source: swan-gtm/gtm-skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.