@grainulation/silo 1.0.0 → 1.0.1

package/packs/hackathon-innovation.json

@@ -0,0 +1,210 @@
+{
+  "name": "Hackathon: Innovation Night",
+  "description": "Creative hackathon category pack for open-ended innovation events. Judging weights: 30% creativity, 25% novelty, 25% evidence quality, 20% feasibility. Rewards divergent thinking and cross-domain synthesis.",
+  "version": "1.0.0",
+  "judging": {
+    "weights": {
+      "creativity": 0.3,
+      "novelty": 0.25,
+      "evidence": 0.25,
+      "feasibility": 0.2
+    },
+    "auto_score_dimensions": {
+      "creativity": {
+        "type_diversity": 0.35,
+        "recommendation_ratio": 0.3,
+        "cross_topic_breadth": 0.35
+      },
+      "novelty": {
+        "challenge_depth": 0.4,
+        "tag_diversity": 0.3,
+        "recommendation_ratio": 0.3
+      },
+      "evidence": {
+        "evidence_tier": 0.5,
+        "corroboration": 0.3,
+        "health": 0.2
+      },
+      "feasibility": {
+        "constraint_satisfaction": 0.4,
+        "estimate_claims_present": 0.3,
+        "risk_claims_present": 0.3
+      }
+    },
+    "human_score_dimensions": [
+      "novelty_of_approach",
+      "synthesis_quality",
+      "practical_creativity"
+    ],
+    "final_blend": {
+      "auto": 0.5,
+      "human": 0.5
+    }
+  },
+  "claims": [
+    {
+      "id": "hin-001",
+      "type": "constraint",
+      "topic": "innovation night format",
+      "content": "Innovation Night is an open-ended creative hackathon format. Teams choose from themed challenge packs or propose their own research question. Output is a 3-minute demo pitch backed by a compiled wheat brief. The format rewards divergent thinking and cross-domain synthesis.",
+      "source": {
+        "origin": "stakeholder",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "stated",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "creative", "format"]
+    },
+    {
+      "id": "hin-002",
+      "type": "constraint",
+      "topic": "creativity scoring",
+      "content": "Creativity is the dominant judging dimension at 30%. Scored across: (1) Type diversity — exploring problems from multiple angles (constraints, risks, estimates, recommendations). (2) Recommendation originality — actionable insights that propose novel combinations. (3) Cross-topic breadth — drawing from diverse knowledge domains to synthesize new understanding.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "creativity", "scoring"]
+    },
+    {
+      "id": "hin-003",
+      "type": "constraint",
+      "topic": "novelty scoring",
+      "content": "Novelty is weighted at 25%. Scored across: (1) Challenge depth — questioning assumptions and resolving conflicting evidence. (2) Tag diversity — exploring across many distinct knowledge areas. (3) Recommendation ratio — generating original actionable insights rather than just collecting facts.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "novelty", "scoring"]
+    },
+    {
+      "id": "hin-004",
+      "type": "constraint",
+      "topic": "evidence scoring",
+      "content": "Evidence quality is weighted at 25%. Even creative ideas need grounding. Scored across: (1) Evidence tier distribution — higher tiers show the team validated their creative ideas. (2) Corroboration — cross-referencing diverse sources is a hallmark of innovative research. (3) Health — clean compilation despite creative exploration.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "evidence", "scoring"]
+    },
+    {
+      "id": "hin-005",
+      "type": "constraint",
+      "topic": "feasibility scoring",
+      "content": "Feasibility is weighted at 20%. Creative ideas must be implementable. Scored across: (1) Constraint satisfaction — does the sprint acknowledge real-world constraints? (2) Estimate claims — are there cost, timeline, or effort projections? (3) Risk claims — has the team considered what could go wrong?",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "feasibility", "scoring"]
+    },
+    {
+      "id": "hin-006",
+      "type": "recommendation",
+      "topic": "themed challenge packs",
+      "content": "Innovation Night should ship with themed creative challenge packs: (1) 'Moonshot' — solve a problem that seems impossible. (2) 'Cross-pollinate' — apply techniques from an unrelated domain. (3) 'Reverse engineer' — start with a desired outcome and work backward. (4) 'Constraint art' — maximum creativity under tight constraints. Themed challenges with creative constraints produce higher quality output than 'build anything'.",
+      "source": { "origin": "research", "artifact": null, "connector": null },
+      "evidence": "web",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "themes", "challenge-packs"]
+    },
+    {
+      "id": "hin-007",
+      "type": "recommendation",
+      "topic": "hybrid voting",
+      "content": "Innovation Night uses hybrid voting: peer voting for 'People's Choice' plus judge panel for category awards. Peer voting increases engagement among non-participants. Categories: Overall Popular Vote, Most Creative, Best Use of AI, Most Likely to Be Implemented.",
+      "source": {
+        "origin": "research",
+        "artifact": "https://info.devpost.com/blog/private-hackathon-voting",
+        "connector": null
+      },
+      "evidence": "web",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "voting", "peer-voting"]
+    },
+    {
+      "id": "hin-008",
+      "type": "risk",
+      "topic": "inclusion and accessibility",
+      "content": "Risk: evening/weekend hackathons exclude parents, caregivers, and anyone with work-life boundaries. Innovation Night template should explicitly offer async submission as an alternative, or run during protected work hours. Non-technical participants must not be disadvantaged — judge on insight quality, not tool proficiency.",
+      "source": { "origin": "research", "artifact": null, "connector": null },
+      "evidence": "web",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "inclusion", "risk"]
+    },
+    {
+      "id": "hin-009",
+      "type": "recommendation",
+      "topic": "demo format",
+      "content": "Standard demo format: 3-minute pitch covering (1) the problem — why does this matter? (2) the approach — what creative angle did you take? (3) the findings — what did you discover? (4) the evidence — how do you know? Demo-first, slides optional. The compiled wheat brief is the supporting artifact.",
+      "source": { "origin": "research", "artifact": null, "connector": null },
+      "evidence": "web",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "demo", "presentation"]
+    },
+    {
+      "id": "hin-010",
+      "type": "recommendation",
+      "topic": "human judge criteria",
+      "content": "Human judges for Innovation Night score on: (1) Novelty of approach — did the team explore an unexpected angle? (2) Synthesis quality — are diverse findings woven into coherent insight? (3) Practical creativity — are recommendations actionable and non-obvious?",
+      "source": { "origin": "research", "artifact": null, "connector": null },
+      "evidence": "stated",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-03-21T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation", "human-judging", "criteria"]
+    }
+  ]
+}

package/packs/hackathon-most-innovative.json

@@ -0,0 +1,179 @@
+{
+  "name": "Hackathon: Most Innovative",
+  "description": "Scoring rubric and seed claims for the 'Most Innovative' hackathon category. Weights type diversity and recommendation ratio heavily — rewards creative, wide-ranging research with novel insights.",
+  "version": "1.0.0",
+  "claims": [
+    {
+      "id": "hinn-001",
+      "type": "constraint",
+      "topic": "type diversity scoring",
+      "content": "Type diversity sub-score = (distinct_types / 6) * 100. Weight: 30% — the dominant factor. Innovative sprints explore problems from multiple angles: constraints, risks, estimates, recommendations, not just facts.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "type-diversity", "most-innovative"]
+    },
+    {
+      "id": "hinn-002",
+      "type": "constraint",
+      "topic": "recommendation ratio scoring",
+      "content": "Recommendation ratio sub-score = (recommendation_claims / total_claims) * 100. Weight: 25%. Innovative sprints produce actionable insights, not just observations. High recommendation ratio signals original thinking.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "recommendations", "most-innovative"]
+    },
+    {
+      "id": "hinn-003",
+      "type": "constraint",
+      "topic": "evidence tier scoring",
+      "content": "Evidence tier sub-score uses weighted sum normalized to 0-100. Weight: 15%. Even innovative ideas need some grounding — but less weight than rigor category.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "evidence-tier", "most-innovative"]
+    },
+    {
+      "id": "hinn-004",
+      "type": "constraint",
+      "topic": "challenge depth scoring",
+      "content": "Challenge depth sub-score = (challenge_claims + resolved_conflicts) / total_claims * 100. Weight: 15%. Innovation often emerges from challenging assumptions and resolving conflicting viewpoints.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "challenge-depth", "most-innovative"]
+    },
+    {
+      "id": "hinn-005",
+      "type": "constraint",
+      "topic": "corroboration scoring",
+      "content": "Corroboration sub-score = witnessed_claims / total_claims * 100. Weight: 15%. Cross-referencing diverse sources is a hallmark of innovative research that synthesizes across domains.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "corroboration", "most-innovative"]
+    },
+    {
+      "id": "hinn-006",
+      "type": "recommendation",
+      "topic": "innovation signals",
+      "content": "Signals of innovation in a wheat sprint: (1) cross-domain tag diversity, (2) claims that bridge multiple topics, (3) recommendations that propose novel combinations, (4) challenges to conventional wisdom with evidence.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "stated",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "innovation-signals", "most-innovative"]
+    },
+    {
+      "id": "hinn-007",
+      "type": "risk",
+      "topic": "diversity gaming",
+      "content": "Teams may add shallow claims of all 6 types just to maximize diversity score. Counter: human judges in final round assess whether type usage is genuine (does the 'risk' claim identify a real risk?).",
+      "source": { "origin": "research", "artifact": null, "connector": null },
+      "evidence": "stated",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "gaming", "type-diversity", "risk"]
+    },
+    {
+      "id": "hinn-008",
+      "type": "factual",
+      "topic": "pointwise rubric evaluation",
+      "content": "Research on automated evaluation shows category-specific rubrics outperform generic rubrics. Each hackathon category needs its own weighted rubric — 'Most Innovative' values different signals than 'Most Rigorous'.",
+      "source": {
+        "origin": "research",
+        "artifact": "https://arxiv.org/html/2503.23989v1",
+        "connector": null
+      },
+      "evidence": "web",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "rubric", "evaluation", "best-practice"]
+    },
+    {
+      "id": "hinn-009",
+      "type": "recommendation",
+      "topic": "non-technical fairness",
+      "content": "Non-technical participants must not be disadvantaged. Guided mode + smart defaults normalize quality. Judge on output quality — insight depth and novelty — not tool proficiency.",
+      "source": {
+        "origin": "stakeholder",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "stated",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "fairness", "non-technical", "most-innovative"]
+    },
+    {
+      "id": "hinn-010",
+      "type": "recommendation",
+      "topic": "human judge criteria",
+      "content": "Human judges for 'Most Innovative' score on: (1) novelty of approach — did the team explore an unexpected angle? (2) synthesis quality — are diverse findings woven into coherent insight? (3) practical creativity — are recommendations actionable and non-obvious?",
+      "source": { "origin": "research", "artifact": null, "connector": null },
+      "evidence": "stated",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "human-judging", "criteria", "most-innovative"]
+    }
+  ]
+}

package/packs/hackathon-most-rigorous.json

@@ -0,0 +1,179 @@
+{
+  "name": "Hackathon: Most Rigorous Research",
+  "description": "Scoring rubric and seed claims for the 'Most Rigorous Research' hackathon category. Weights evidence tiers heavily, rewards challenge depth and corroboration over volume.",
+  "version": "1.0.0",
+  "claims": [
+    {
+      "id": "hrig-001",
+      "type": "constraint",
+      "topic": "evidence tier scoring",
+      "content": "Evidence tier sub-score uses weighted sum: production=5, tested=4, documented=3, web=2, stated=1, normalized to 0-100. This is the dominant factor (35% weight) in the Most Rigorous category.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "evidence-tier", "most-rigorous"]
+    },
+    {
+      "id": "hrig-002",
+      "type": "constraint",
+      "topic": "challenge depth scoring",
+      "content": "Challenge depth sub-score = (challenge_claims + resolved_conflicts) / total_claims * 100. Weight: 25%. Rewards sprints that actively question and resolve competing claims rather than building one-sided arguments.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "challenge-depth", "most-rigorous"]
+    },
+    {
+      "id": "hrig-003",
+      "type": "constraint",
+      "topic": "corroboration scoring",
+      "content": "Corroboration sub-score = witnessed_claims / total_claims * 100. Weight: 20%. Claims backed by external sources (web evidence, documented references) score higher than unsupported assertions.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "corroboration", "most-rigorous"]
+    },
+    {
+      "id": "hrig-004",
+      "type": "constraint",
+      "topic": "health scoring",
+      "content": "Health sub-score = max(0, (1 - warnings/total_claims) * 100). Weight: 10%. Penalizes sprints with compiler warnings — low-evidence claims, missing fields, unresolved conflicts.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "health", "most-rigorous"]
+    },
+    {
+      "id": "hrig-005",
+      "type": "constraint",
+      "topic": "type diversity scoring",
+      "content": "Type diversity sub-score = (distinct_types / 6) * 100. Weight: 10%. Ensures sprints use the full taxonomy (constraint, factual, estimate, risk, recommendation, feedback) rather than only one type.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "scoring", "type-diversity", "most-rigorous"]
+    },
+    {
+      "id": "hrig-006",
+      "type": "recommendation",
+      "topic": "anti-gaming logarithmic scaling",
+      "content": "Apply logarithmic scaling to claim counts: score = log(claim_count) not linear. This prevents gaming through claim spam — 100 weak claims score only marginally better than 10 strong ones.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "anti-gaming", "logarithmic", "most-rigorous"]
+    },
+    {
+      "id": "hrig-007",
+      "type": "risk",
+      "topic": "Goodhart's Law gaming",
+      "content": "If teams know the exact scoring weights, they optimize for metrics over substance (Goodhart's Law). Keep exact weights secret; publish only the dimensions. Rotate weights per hackathon event.",
+      "source": {
+        "origin": "research",
+        "artifact": "https://en.wikipedia.org/wiki/Goodhart's_law",
+        "connector": null
+      },
+      "evidence": "web",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "gaming", "goodhart", "risk"]
+    },
+    {
+      "id": "hrig-008",
+      "type": "factual",
+      "topic": "benchmark thresholds",
+      "content": "A well-designed rigorous sprint should achieve: evidence_tier_score >= 60, type_diversity >= 83 (5/6 types), challenge_depth >= 15%, corroboration >= 20%, health >= 80. These serve as calibration benchmarks.",
+      "source": { "origin": "research", "artifact": null, "connector": null },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "benchmarks", "calibration", "most-rigorous"]
+    },
+    {
+      "id": "hrig-009",
+      "type": "recommendation",
+      "topic": "hybrid judging model",
+      "content": "Two-stage judging: auto-score (compiler metrics) acts as qualifier — top N advance. Final round = 50% auto-score + 50% human panel scoring insight quality, narrative clarity, and practical applicability.",
+      "source": { "origin": "research", "artifact": null, "connector": null },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "research",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "hybrid-judging", "qualifier", "most-rigorous"]
+    },
+    {
+      "id": "hrig-010",
+      "type": "constraint",
+      "topic": "evidence tier floor",
+      "content": "Claims below 'web' evidence tier count at 50% weight. This prevents gaming by adding many low-effort 'stated' claims. Encourages teams to back assertions with real sources.",
+      "source": {
+        "origin": "best-practice",
+        "artifact": null,
+        "connector": null
+      },
+      "evidence": "documented",
+      "status": "active",
+      "phase_added": "define",
+      "timestamp": "2026-01-01T00:00:00.000Z",
+      "conflicts_with": [],
+      "resolved_by": null,
+      "tags": ["hackathon", "anti-gaming", "evidence-floor", "most-rigorous"]
+    }
+  ]
+}