specsmd 0.0.0-dev.82 → 0.0.0-dev.84

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -21,10 +21,14 @@ Generate genuinely diverse, surprising ideas through rapid batches. AI is the cr
21
21
  <llm critical="true">
22
22
  <mandate>NEVER ask setup questions — generate ideas immediately on receiving a topic</mandate>
23
23
  <mandate>NEVER name techniques to the user — use them internally, present only the ideas</mandate>
24
- <mandate>NEVER generate 2+ ideas from the same domain in a batch — enforce anti-bias</mandate>
25
24
  <mandate>ALWAYS use the deep thinking protocol before generating each batch</mandate>
26
25
  <mandate>ALWAYS present ideas as vivid, specific, memorable concepts — not abstract descriptions</mandate>
27
26
  <mandate>NEVER more than 2 questions in a row — always generate something creative first</mandate>
27
+ <mandate>ANTI-BIAS — MATCH DIVERSITY TO TOPIC TYPE:
28
+ - For OPEN/CREATIVE topics (product ideas, business concepts, experiences): use the domain wheel aggressively — 3+ different domains per batch, cross-domain metaphors welcome
29
+ - For TECHNICAL/ARCHITECTURAL topics (system design, integration decisions, API design): diversity means different PERSPECTIVES and ANGLES on the same technical space — architecture, developer experience, tradeoffs, migration strategy, patterns from other systems. Do NOT force nature/art/games metaphors onto technical decisions. The ideas should feel like they came from 5 different senior engineers, not 5 different departments.
30
+ - When in doubt, match the user's register. If they're talking code, give them code-level ideas.
31
+ </mandate>
28
32
  </llm>
29
33
 
30
34
  <protocols>
@@ -64,23 +68,28 @@ Generate genuinely diverse, surprising ideas through rapid batches. AI is the cr
64
68
  <action>Execute the deep thinking protocol (see shared/protocols/deep-thinking.md)</action>
65
69
  <action>Perform 6-step reasoning chain:</action>
66
70
  <reasoning>
67
- [Think 1 — Domain Check]: Which domains from the domain wheel are underexplored? Select 3+ for this batch.
68
- [Think 2 — Raw Concepts]: Generate raw ideas — one per selected domain. Use a technique internally (SCAMPER, analogy, inversion, etc.).
71
+ [Think 1 — Angle Check]: What topic type is this? For creative/open topics, pick 3+ domain wheel sectors. For technical topics, pick 3+ different perspectives (architecture, DX, tradeoffs, patterns, prior art, migration, cost). Either way, ensure the batch won't feel repetitive.
72
+ [Think 2 — Raw Concepts]: Generate raw ideas — one per selected angle. Use techniques internally (SCAMPER, analogy, inversion, first-principles, etc.) but keep ideas grounded in the user's actual domain.
69
73
  [Think 3 — Novelty Filter]: Are these genuinely new or just repackaged versions of obvious ideas? Replace any that feel generic.
70
74
  [Think 4 — Cross-pollinate]: Can any ideas combine with user's stated favorites? Create unexpected connections.
71
- [Think 5 — Provocation]: If provocation is due (every 15 ideas), inject one deliberately absurd idea. Otherwise, ensure at least one idea challenges assumptions.
72
- [Think 6 — Polish]: Make each idea vivid, specific, and memorable. Not "use gamification" but "a quest system where each completed task unlocks a new tool."
75
+ [Think 5 — Provocation]: If provocation is due (every 15 ideas), inject one deliberately unconventional idea. Otherwise, ensure at least one idea challenges assumptions.
76
+ [Think 6 — Polish]: Make each idea vivid, specific, and memorable. Match the user's register if they're talking about code, give them code-level specificity, not abstract metaphors.
73
77
  </reasoning>
74
78
  </step>
75
79
 
76
80
  <step n="3" title="Generate Spark Batch">
77
81
  <action>Present batch of 5 ideas, each with:</action>
78
82
  <format>
79
- **#{number}** — {vivid idea title}
83
+ **S{batch}-{number}** — {vivid idea title}
80
84
  {2-3 sentence description — specific, concrete, imaginative}
81
85
  </format>
82
86
  <action>Number ideas sequentially across batches (S1-1, S1-2... S2-1, S2-2...)</action>
83
- <anti_bias>Verify: batch spans 3+ domains, no 2 consecutive ideas from same domain</anti_bias>
87
+ <anti_bias>
88
+ BEFORE presenting the batch, verify diversity:
89
+ - For creative/open topics: 3+ different domain wheel sectors
90
+ - For technical topics: 3+ different perspectives (e.g., architecture, DX, tradeoffs, patterns, migration)
91
+ - No 2 consecutive ideas making the same argument from the same angle
92
+ </anti_bias>
84
93
  </step>
85
94
 
86
95
  <step n="4" title="Collect Reactions">
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "specsmd",
3
- "version": "0.0.0-dev.82",
3
+ "version": "0.0.0-dev.84",
4
4
  "description": "Multi-agent orchestration system for AI-native software development. Delivers AI-DLC, Agile, and custom SDLC flows as markdown-based agent systems.",
5
5
  "main": "lib/installer.js",
6
6
  "bin": {