decision
Skill0 starsChanged 4 months agoMITView on GitHub
Choose build, reject or wait for each opportunity.
19 other projects use a version of this.
Adopted by
Most recently taken up by
- timursafiullin/aodtheir own version, first seen Nov 15, 2026
- Renozoic-Foundry/forge-publictheir own version, first seen Nov 15, 2026
- sohaibt/product-modetheir own version, first seen Nov 8, 2026
- lulufoo/lulu-workflowtheir own version, first seen Nov 6, 2026
- qbicus/skillstheir own version, first seen Oct 25, 2026
- emiljuchnikowski/ai-os-templatetheir own version, first seen Oct 20, 2026
Used with
Projects with it also use explore, brief and 5 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
genomewalkerSkillexplore
RLM-style recursive memory exploration - dynamically navigate the memory graph
Also in 3 projects that use this
AnthropicSkillbrief
Generate contextual briefings for legal work — daily summary, topic research, or incident response. Use when starting your day and need a scan of legal-relevant items across email, calendar, and contracts, when researching a specific legal question across internal sources, or when a developing situation (data breach, litigation threat, regulatory inquiry) needs rapid context.
Also in 3 projects that use this
BlockSkillplan
Draft a TDD, PRD, or lightweight plan in the conversation without committing it.
Also in 4 projects that use this
SentrySkillspec
Produce a verified spec — problem, desired outcome, and acceptance criteria — before building. Use when given a GitHub/Linear issue link, when starting from a fuzzy in-conversation idea, or asked to "figure out what to do" / "scope this" / "what should I build" — whenever the *what* isn't yet pinned down. Grills the ambiguities, checks the premise against the code, and produces acceptance criteria you sign off on — then hands to design-first.
Also in 3 projects that use this
OpenAISkillaudit
Audit or critique a product flow, journey, workflow, funnel, onboarding path, checkout path, settings path, screen, or multi-step product experience by capturing screenshots first, then reporting UX, design, and accessibility findings inline from that evidence. Use Figma only when the user explicitly asks for a board. Use when the user asks to audit, review, critique, inspect, assess, analyze, evaluate, or give feedback on a product experience.
Also in 3 projects that use this
SentrySkillreview
Fixture review skill.
Also in 4 projects that use this
Other versions19 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 19
- sohaibt/product-mode★ 171same name
- aleksander-dytko/ai-pm-workspace★ 34same name
- chrisalehman/bionic★ 9same name
- mhxie/Atelier★ 5same name
- lulufoo/lulu-workflow★ 3same name
- madebypx/PXOS★ 3same name
- Jsnnmsc/10000x-engineer★ 2same name
- roethlar/AgentGovernanceBootstrap★ 2same name
- Tamircohen28/tamirs-superpowers★ 2same name
- yacb2/aidex★ 2same name
9 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository11 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- analyze-competitionSkill
- analyze-marketSkill
- cluster-painsSkill
- collect-evidenceSkill
- compose-reportSkill
- contradiction-analysisSkill
- discover-opportunitiesSkill
- harvest-leadsSkill
- interview-planSkill
- research-briefSkill
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 1 is distinctive, and 4 are very specific to this project.
0
Common
Most of the 18 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 61
Every stage before this one was forbidden from netting its findings against anything.
- Very specific: line 78
A crowded field is not by itself a reject: Read competitors together with switching_costs and substitutes.
- Very specific: line 85
differentiationis a claim, not a finding: Treat an entry there as something still to be validated unless something in the evidence … - Very specific: line 98
Read the basis and decide what the gap actually is:
- Very specific: line 133
If a contradiction analysis found nothing, note that you saw it — and note whether its searched_for list was thorough enough for the …
“Common” is measured: kept by at least 6 in 10 of the 18 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Decide
Rule on each opportunity below — build, reject or wait — with reasoning explicit enough that someone can re-judge the call later and see exactly where they disagree with you.
The question
${question}The opportunities
${opportunities}The market analyses
${market}The competition analyses
${competition}The contradiction analyses
${contradictions}One verdict per opportunity
Exactly one, linked to the opportunity by its real id. There is no "build with
reservations": that is build, plus a next validation step that says what you are watching.
- build — pursue it now. The case holds and the remaining risks are ones you are knowingly accepting.
- reject — do not pursue, and do not expect that to change. A judgement about the idea: the pain is not worth solving, or this is not the way to solve it.
- wait — plausible, but something has to change or be learned first. Name what, and
name what would tell you it has happened. A
waitwith no trigger is avoidance wearing a verdict.
The distinction most often collapsed is reject versus wait. Rejecting because you do
not know enough closes a question nobody actually examined; waiting on an idea that is
plainly bad burns a research cycle. Be clear which one you mean. And do not use wait as
the safe middle: if every opportunity comes back wait, you have decided nothing.
This is the only stage that weighs
Every stage before this one was forbidden from netting its findings against anything. The market analysis made the case for and did not discount it; the competition analysis described the field and did not conclude from it; the contradiction analysis returned findings and no verdict. The trade-off is unresolved on purpose, and it is yours. Say plainly what you traded against what — the weighing has to be visible in your reasons, not just implied by the verdict you landed on.
How to weigh the four readings
The dimensions are not interchangeable. They do not add up to a score, and a strong reading on one does not buy down a weak reading on another.
- A strong market case does not rescue an opportunity with a blocking contradiction. A
large, well-derived market makes a blocked opportunity a bigger miss, not a smaller
risk. If a finding is graded
blocking, the size of the prize is not the answer to it — either a reason states plainly why that finding does not bind this decision, or the verdict iswaitwith the finding as the trigger. - A crowded field is not by itself a reject. Read
competitorstogether withswitching_costsandsubstitutes. Many incumbents with low switching costs means customers can move — the field is contested but not closed, and the honest verdict is usuallywaiton evidence that a wedge exists, notreject. High switching costs, or amoatsentry that a new entrant has no route around, is what turns a crowded field into a genuine reject. Note also thatsubstitutesare usually the real competition: a spreadsheet that is good enough is a harder opponent than a funded company. differentiationis a claim, not a finding. Treat an entry there as something still to be validated unless something in the evidence actually supports it.
An unsized market is missing information, not a negative signal
A SizeEstimate can legitimately carry no number. Its basis is mandatory either way: with
an amount it is the derivation, without one it is why the market could not be sized. The
previous stage was built to make that refusal expressible so that nobody would have to
invent a figure.
So do not read an absent SAM or SOM as a small market. Treating "we could not size
this" as evidence against would punish exactly the honesty the stage was designed for, and
it would teach the pipeline that a fabricated number scores better than an admitted gap.
Read the basis and decide what the gap actually is:
- Unsized because the buyer segment is not yet identifiable → that is your
biggest unknownand probably awaitwith sizing as the next validation step. - Unsized because the figure exists but was not to hand → it is a cheap next validation step and should barely move the verdict.
- Sized with a
basisthat is a chain of guesses → weaker than it looks. A number with a thin derivation deserves less weight than an honest refusal.
The same reading applies to pricing, to assumptions, and to unknowns — an empty
unknowns list is rarely honest, and is itself worth noticing. And check the buyer
against the budget_owner: when they differ, someone has to be persuaded who does not feel
the pain, and that is a real cost the size figure does not capture.
Reasons must be traceable
Each reason points at something in the artifacts you were given — an opportunity's why
now, a size estimate that could not be established, a switching cost, a specific piece of
counter-evidence — and cites it by real id where it has one. A reason resting on a fact
that appears in none of these artifacts is not a reason; it is a preference.
Give the two or three things that actually drove the call, strongest first, not everything true about the opportunity. Include what argued the other way and why it lost. A list of only supporting points is advocacy, and it gives a later reader nothing to check.
The counter-evidence has to show up
The contradiction analyses were produced deliberately, without weighing, so that the weighing happens here and is visible.
- A
blockingfinding pushes hard away frombuild. Building over one is permitted, but only if a reason states plainly why that finding does not bind this decision. Otherwise the honest verdict iswait. materialfindings do not block, but should appear somewhere: in a reason, in the next validation step, or in an explicit statement that the call holds either way.- If a contradiction analysis found nothing, note that you saw it — and note whether its
searched_forlist was thorough enough for the emptiness to mean anything. An empty result from a shallow search is not reassurance.
Confidence, unknown, next step
Required for every verdict, including reject, and including a build you are sure of.
- Confidence grades this judgement, not the idea's attractiveness. Anchor it to how
much the artifacts actually support the call. Thin, sparse evidence with a clear
direction is a low-confidence
build, and saying so is what makes the call reversible. Four readings that agree do not raise confidence if all four rest on the same one piece of evidence — check whether they are independent before you let them accumulate. - Biggest unknown is required even when you are confident — especially then. A confident verdict that cannot name what it is still most likely to be wrong about has stopped reasoning. Name the specific thing, not "market dynamics".
- Next validation step is the single cheapest action that would move the confidence:
who to talk to, what to look up, what to try. Days, not quarters. It applies to
rejectas much asbuild— what would make you revisit this, and what do you tell whoever asked the question?
Failure modes
- A verdict that does not follow from its own stated reasons.
- Netting the four readings into an implied score, or letting market size answer an objection that is not about size.
- Reading an unsized market, or an honest
basis, as evidence against. - Rejecting on the mere presence of competitors, without reference to switching costs, substitutes or moats.
waitused to avoid deciding, or with no trigger attached.- Building on everything, or rejecting everything, which is the same failure twice.
- Reasoning that appeals to facts none of the supplied artifacts contain.
- Silence about the strongest argument against the verdict you chose.
- Confidence set by how good the idea sounds rather than by how much stands behind it.
- A biggest unknown so vague that no evidence could ever resolve it.
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.

