launch-retrospective
Skill417 starsChanged 3 months agoMITView on GitHub
Use after a social launch, campaign, creator collaboration, or distribution test has enough evidence to decide what should be repeated, changed, or stopped. Trigger phrases include "run a launch retro", "what did we learn from this campaign?", "postmortem this launch", "should we repeat this?", and "turn these results into next steps". Do not use this to assign blame, invent missing data, or write a live performance dashboard; use social-analytics-report for the in-flight metric read and campaign-brief to plan the next campaign.
In this repository54 other files it gives its agents.
- ab-hook-testerSkill
- agent-reachSkill
- ai-brainSkill
- analytics-readoutSkill
- audience-personasSkill
- best-time-schedulerSkill
- bio-profile-optimizerSkill
- b-roll-shot-listSkill
- campaign-briefSkill
- caption-and-hashtagsSkill
- carousel-builderSkill
- collab-outreachSkill
- comment-miningSkill
- comment-responderSkill
- community-engagement-planSkill
- competitor-positioningSkill
- competitor-teardownSkill
- content-auditSkill
- content-calendarSkill
- content-factorySkill
- content-pillar-builderSkill
- content-repurposerSkill
- content-researchSkill
- cover-thumbnail-briefSkill
- creator-outreachSkill
- cross-platform-distributionSkill
- cross-platform-reformatterSkill
- cta-writerSkill
- dm-script-writerSkill
- going-viralSkill
- hashtag-keyword-researchSkill
- hook-miningSkill
- humanizerSkill
- influencer-partnershipsSkill
- linkedin-thought-leadershipSkill
- on-screen-text-writerSkill
- paid-social-briefSkill
- positioning-auditSkill
- profile-conversion-auditSkill
- reel-analyzerSkill
- reel-builderSkill
- reel-scripterSkill
- series-plannerSkill
- social-analytics-reportSkill
- social-listeningSkill
- social-proof-miningSkill
- story-sequencerSkill
- trend-spotterSkill
- ugc-creator-briefSkill
- utm-campaign-planSkill
- viral-carouselSkill
- viral-hook-writerSkill
- viralitySkill
- x-thread-strategySkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 16
Use available performance data, qualitative feedback, partner notes, approvals, and execution records.
- Distinctive: line 28
Do not use a retrospective to blame individuals, creators, or customers for an outcome.
- Distinctive: line 29
Never claim causation where the evidence only shows correlation.
- Distinctive: line 31
Do not compare campaigns with different objectives or measurement conditions as if they were identical.
- Distinctive: line 45
Use social-analytics-report for structured performance evidence during the launch.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Launch Retrospective
Turn a completed campaign into durable operational learning: what happened, why it may have happened, and the next smallest decision.
1. Reconstruct the intended plan
Bring together the original objective, audience, approved message, channel roles, budget or partner scope where supplied, timing, measurement plan, and decisions made during execution. Compare what actually shipped with what was planned. Separate a deliberate change from an accidental deviation.
2. Gather the evidence with limits attached
Use available performance data, qualitative feedback, partner notes, approvals, and execution records. State dates, metric definitions, missing data, and material confounders. Keep a timeline so the team does not attribute an outcome to a change that happened after it.
3. Distinguish facts from explanations
List observations first: what was published, what audiences did, and what tracking shows. Then list hypotheses about message, format, distribution, timing, destination, or process. Give each hypothesis a confidence level and the evidence that would strengthen or weaken it.
4. Commit to operational changes
Return a short set of keep, change, test, and stop decisions with owners and due dates. Update reusable inputs such as the campaign brief template, proof library, UTM vocabulary, or creator brief. Leave a clear record of unresolved questions so they are not rewritten as facts next time.
Hard rules
- Do not use a retrospective to blame individuals, creators, or customers for an outcome.
- Never claim causation where the evidence only shows correlation.
- Include negative and inconclusive results; they are part of the learning.
- Do not compare campaigns with different objectives or measurement conditions as if they were identical.
- Separate process failure from message failure before prescribing a fix.
Failure modes
| Failure | Do this instead |
| The retro is a victory lap or blame document | Use a factual timeline, uncertainty, and future-oriented decisions. |
| A single metric explains the whole launch | Combine the original job, data boundary, qualitative evidence, and process record. |
| Lessons stay as vague observations | Turn each learning into a named change, test, or stop decision. |
| The same tracking issue returns next launch | Update the UTM or measurement workflow with a clear owner. |
Where it sits
Use social-analytics-report for structured performance evidence during the launch. Use campaign-brief, paid-social-brief, influencer-partnerships, and utm-campaign-plan as source records. Feed confirmed lessons into the next campaign-brief rather than treating the retro as a standalone document.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.