Judge the decision, not the result.
The tool that reviews something that went wrong, or surprisingly right, and separates the part that was your decision from the part that was luck, so you learn the lesson that survives.
The most expensive mistake after the fact is judging a decision by how it turned out. A deal that closed proves the pitch was right. A launch that flopped proves the idea was wrong. Neither is true. Outcomes are decisions plus luck, and people draw the obvious lesson and bake the luck into their strategy as if it were skill.
What Actually Happened pulls the two apart. It reconstructs what was known at the time of the decision, separates that from what only became visible afterwards, and names the lesson that would still hold if the dice had landed differently.
Learn the right lesson, not the obvious one.
Same lost deal, same five minutes. Here is the post-mortem, run two ways.
You lost a £200k deal you were sure you would win. It went to a competitor in the final week, after four months of positive signals. You want to know what you did wrong.
You ask what went wrong. You get a thoughtful list. Maybe you should have qualified harder, built more champions, competed on price, followed up sooner. All plausible. All retrofitted to the fact that you lost.
So you walk away with a lesson shaped by the outcome, ready to overcorrect on the next deal that looks nothing like this one.
It refuses to judge the decision by the result. First it separates what you knew then from what you learned after:
- →Known then: a parallel procurement process you glossed over
- →Hindsight: the competitor's 40% final-week price cut
- →Hindsight: a new CFO who favoured the incumbent
Then the lesson that survives: when a known parallel process exists, find out who else is bidding by the third meeting. It survives because it would have helped regardless of the price cut. The lesson is NOT compete on price, the cut was luck, and chasing it would win you worse deals.
The cold version gave you a lesson shaped by the loss. This gave you the one that survives the luck.
- ✦Founders reviewing a deal lost, a launch flopped, or a hire that failed
- ✦Operators who know a good outcome can follow a bad decision, and the reverse
- ✦Anyone tempted to rewrite a whole strategy on the back of one vivid result
- →what-actually-happened.zip: a native Claude Skill. Upload to Settings → Skills in Claude, or unzip to
~/.claude/skills/ - →what-actually-happened-skill.md: plain markdown. Works in ChatGPT Custom GPTs, Gemini, Claude Projects, agent system prompts, anywhere
- →A README inside the zip with install steps for Claude, ChatGPT, Gemini, and the API
What it does with a real question.
Last quarter I launched a new service line with a six-week sprint, spent $40,000 on it, and signed only two clients against a target of ten. The market then softened in month two. I am being told it failed because of the economy. I am not sure that is true, and I want to learn the right lesson.
Your doubt is justified, but so is the economy. What you've described is a mix, and the useful work is pulling the two apart.
Start with what you knew when you committed the $40,000. The softening hit in month two, after the spend was already made, so it could not have been weighed in the decision. That part is luck, and no lesson hides in it. But a soft market does not explain 2 of 10. Even a rough economy rarely cuts a pipeline by 80 percent. A miss that size usually means the target itself was never grounded in anything, and that part was in your control.
The question I would press on is where the ten came from. If it came from hope or from a gut read on demand, rather than from conversations with buyers who had signaled they would pay, then the decision was weaker than the economy story lets it be. If you had ten warm prospects in the pipeline before you spent a dollar, the economy explanation gains real weight.
I would not conclude "never launch in uncertain markets" or "this service is dead." Two paying clients proves some demand exists.
Your next move: pull the notes on every prospect you spoke to. Sort them into lost to budget freeze, lost to fit or price, and never really interested. That split tells you whether you had a timing problem or a demand problem, and only the second one should change your strategy.
The full Tool runs the whole procedure and writes the complete deliverable, not this excerpt.
Written by claude-sonnet-5-5 with this Tool loaded, 2026-10-11. Checked automatically for length and for copied framework text.
No subscription · free updates within v1.x
Works as a Claude Skill (.zip) and as a plain .md for ChatGPT, Gemini & every LLM
Delivered to your email within 60 seconds · Secured by Stripe
Price includes UK VAT where applicable
Before you buy.
How is this different from just asking Claude directly?
Asked cold, Claude retrofits a list of lessons to the fact that things went wrong, which is exactly the trap, judging the decision by its outcome. This reconstructs what was knowable at the time, judges the decision on that alone, separates the luck from the skill, and keeps only the lesson that would hold if the result had gone the other way. Sometimes that lesson is nothing to change.
Will this work with ChatGPT, Gemini, and other LLMs?
Yes. You get a native Claude Skill (.zip) that auto-invokes in Claude, plus a plain .md you can paste into a ChatGPT Custom GPT, a Gemini system prompt, or any LLM. The procedure travels with the file.
Can I share this with my team?
It's licensed for your own personal and professional use. The output it produces is yours to circulate. To put the Tool itself in the hands of a whole team, email hello@authority.md about team licensing.
What if it doesn't work for me?
If the file hasn't been delivered, or arrives defective or empty, we refund in full: see our refund policy.
What Actually Happened is a decision-review tool. Not a substitute for an honest team debrief or a real conversation with the other party.