dbs-standard-answer starts with a direct answer to your question, then does two linked pieces of research. Part one reviews the claim for absolutes and causal leaps, maps relevant fields, authors and theories, verifies the main theory’s source and applies it to your situation. Part two turns the problem into a structural fingerprint and compares historical cases with the same structure: nearest successes, cross-field repeats, failures and counterexamples.
The result is graded as mature consensus, conditional answer or no unified answer, and is written as when X holds, the repeated approach is Y because Z, and it fails when W. Short questions get a short merged answer, and video, course or slide versions are added after the research when requested.
What You Get
- A direct answer followed by a claim review (keep, narrow, rebuild or drop)
- A table of fields, authors, theories and mechanisms with verification status
- Four to six historical cases with success, failure and counterexample roles, and a similarity matrix
- A conditional standard answer with its cause, its failure boundary and a list of open facts
- Optional short-video prototypes, slide or course versions built on the research
Common Use Cases
- Asking whether a classic solution exists for a recurring business or career problem
- Finding credible theory behind a method before turning it into a course or video
- Comparing how similar founders or organisations handled the same bind
What Makes It Unique
It joins theory research and historical case comparison in one flow, and the mechanisms found in part one decide which cases enter part two. It grades its own conclusion, separates verified facts from interpretation and inference, and states where the answer breaks. Quotes found through secondary sources are labelled as such, and unverifiable anecdotes are not written as fact.
Under The Hood
The skill is one long instruction file with no scripts. It works from your material and the model’s own knowledge by default. It may search the web only when an external fact would change the conclusion or you ask for sources, and it stops after two failed attempts to verify a fact. It does not write files unless you ask for a deliverable, and it prefers original works, papers and official archives over encyclopedias when it does check.