Four spec-driven tools, one sentence, 6x gap from idea to agent-ready plan
SSShken · reddit · 2026-09-11
Testing four spec-driven tools (OpenSpec, Spec Kit, BMAD, Kiro) with one identical sentence on the same machine: actions to a full task list ranged from 14 (Spec Kit) to 54 (Kiro, 32 of which were clicking Allow dialogs), a 6x gap between "I have an idea" and "my agent can start". All four recommended different stacks, none offered task editing, all added an unwanted delete function, and only Kiro required an account and admin rights.
Related event: Same Prompt, Four AI Spec Tools: Planning Steps Differ by 6x(2 posts)→
More from coding & agent
- HABIT-Bench: Best agent memory system scores just 33.61% at inferring user habits — SJTU · 2026-10-28
- Memory Tax study: dedicated agent memory systems often lose to naive RAG — UIUC-CS · 2026-10-28
- Regehr's 'Piano Test': would you trust formally verified code with your life? — moyix · 2026-10-28
- Vibe-coded 'AI 浮生记' game satirizes AI startup life with compute black-market trading — vista8 · 2026-10-28
- Cursor ships Annotate in design mode: agents take visual prompts via on-page markup — soleio · 2026-10-28
- The legendary fake Go generics: Unicode Canadian Syllabics as angle brackets — rickasaurus · 2026-10-28