A curated corpus of sample inputs and artifacts for testing and demonstrating BMad Method skills.
Around 100 entries spanning the full BMad funnel: one-line intents, braindumps, transcripts, threads, BRDs, briefs, PRDs, PRFAQs, OSS RFCs, JTBD specs, and pre-written Spec kernels. Each artifact is hand-shaped to test a specific skill, edge case, or product-thinking pattern.
See index.md for the full catalog, per-skill testing recommendations, and the experiments this corpus enables.
intents/— minimum-density inputs (one-liners, tweets, SMS threads, bullet lists, README stubs)braindumps/— founder-energy raw inputs (rants, urgency memos, stealth blogs, post-coffee notes)transcripts/— verbal-source captures (voice memos, podcasts, interviews, walking meetings)threads/— multi-voice conversations (Slack, email, X, GitHub issues)research/— research outputs (customer-interview synthesis, market analysis)bmad-spec-kernel/— pre-written five-field Spec kernels (clean exemplars and violation studies)brds/— Business Requirements Documents (corporate "the system shall" style)briefs/— product briefs (one altitude above PRDs; includes brief→PRD pairs)prds/— full PRDs, PRFAQs, OSS RFCs, JTBD specsarchitectures-solution-designs/,projects/,other/— placeholders for downstream artifacts
Point a BMad skill at any entry. The catalog in index.md maps each skill to the inputs that exercise it.
MIT. See LICENSE.