- Home
- Skills
- Product Management
- Feature Launch Playbook
More screenshots
Works with the AI Agents you already use
Feature Launch Playbook
Launch features so they land: five go/no-go gates covering the story, org readiness, safe rollout with a kill switch, sequenced comms, and a measured verdict.
$5
Feature Launch Playbook
Example session with this skill installed
We're a 14-person B2B SaaS. Next month we ship recurring invoices for our billing tool. Plan the launch.
- Read your context and instructions
- Compiled the feature launch playbook
Tier: Standard, leaning Major. Recurring invoices change what users can do and give sales new words — but there's no pricing change, so we run Standard with the Major-level internal brief.
Gate 1 — the one-page story
- Who: finance leads at 10–50-person firms who bill the same clients monthly.
- Their complaint: 'I rebuild the same invoices by hand on the 1st and chase the ones I forgot.'
- Now: a calendar reminder, a spreadsheet of who's due, and manual re-entry.
- Promise: 'Set up an invoice once; it sends itself on schedule, every month.'
- Proof: recurring schedules per client; automatic send with the existing payment links; a missed-payment digest instead of silent failures.
- Not: not accounting software — no ledger, no tax filing.
Gate 3 — rollout
Gradual ramp to existing paying accounts: 5% → 25% → 100% over two weeks, new signups last. Rollback rules, written now: error rate on invoice sends above 1% for 5 minutes → pause and revert; feature-tagged tickets above 20 in a day → pause ramp, escalate to the named owner.
Gate 4 — sequence
T-7d: support trained, FAQ live with the known limitation (no mid-cycle proration in v1). T-2d: CS emails the 30 accounts who requested this personally. Launch day, after the 25% health check passes: in-app banner + email to active admins. Blog post and release notes same day, manually triggered — nothing auto-fires.
Gate 5 — verdict, set for 4 weeks out
Hypotheses: 20% of eligible accounts create a schedule in month one; manual-invoice support tickets drop 15%. If adoption misses, diagnose in order — heard about it? tried and bounced? tried and shrugged? — before changing anything.
Connects securely to your tools. The creator never sees your data.
About this skill
Engineering finishing a feature and the company getting value from it are two different events, and the gap between them is where most launches die. Sales cannot pitch what they have not been told about. Support cannot help with what they have not been trained on. Customers cannot use what they never heard of. The Feature Launch Playbook closes that gap by running every launch as five go/no-go gates, each with a concrete output and a bar it must clear.
Gate 0 sizes the launch honestly — Quiet, Standard, or Major — so big launches get the full treatment and small ones do not get a parade. Gate 1 produces the one-page story: who it is for (specific enough to picture a person), their complaint in their words, what they do now, the promise, the proof, and what the feature is not. Gate 2 readies the org with a timed internal brief, a sales battlecard and demo, support training with a living FAQ, and a named executive sponsor. Gate 3 picks a rollout pattern matched to blast radius and confidence, starts monitoring before the rollout does, and writes the kill-switch rules in advance — so a mid-launch problem is executed on, not debated. Gate 4 sequences comms so public announcements fire only after a rollout health check, never on a timer into a paused rollout. Gate 5 delivers the verdict weeks later, on a stable trend: did they come, did they stay, did the number move, did anything break — with a diagnosis framework when adoption misses.
The playbook includes the full launch-plan template, a worked example, a launch-readiness score for the review meeting, edge cases (rollback during the announcement window, leaks, a sponsor going dark), and a twelve-entry rogues' gallery of the ways launches die — every one of them a skipped gate. It scales from a 14-person SaaS to a product-led launch with no sales team at all.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Product Management.
- It governs the launch, not the feature: a weak feature with a flawless launch still fails Gate 5's verdict, and the playbook will say so.
- Quiet releases do not need it — a changelog line and monitoring are the honest launch there.
- The verdict needs weeks of data by design; teams that must declare success on launch day will find Gate 5 unsatisfying on purpose.
- Battlecard and briefing formats assume a B2B or hybrid motion; pure consumer launches compress those sections, as the skill notes.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 24 days ago
Permissions
No special permissions declared or detected