# United We Transform: full context for agents

## What this is
A neutral public evidence atlas for gatherings over 25 people, built from
23,624 publicly posted event agendas. Every agenda is scored 0-100 (the
Gathering Effectiveness Score, GES) by a deterministic rubric (ges_agenda_judgment_v2_0)
across eight pillars. The rubric is fully published; scoring has no black
box and runs the same way client-side at https://unitedwetransform.com/grader/ as it does server-
side across the corpus. A 1,000-event gold set is hand-verifiable and is the
only part of the corpus treated as verified; everything else is labeled
provisional, scored from the published agenda only.

## The two-class honesty rule
Gold-verified events (1,000): evidence is hand-checkable; these appear in the
leaderboard and search index (both are gold-only by design, not a temporary
limitation). Provisional events (the rest, 23,624 total): scored from the
published agenda alone; negative signals mean "not visible in the public
agenda," never "did not happen." Letter grades are withheld corpus-wide until
human calibration (a blind hand-scored validation set) confirms the bands;
until then, use the numeric ges field, never an inferred letter. Every score
object carries verification_tier, per-pillar evidence_provenance
(source-backed vs inferred), and rating_basis (rubric + model layers + human
calibration status).

## The scoring formula (from the published rubric)
Step 1: agenda_design_potential = clamp((sum of the 7 mechanism pillar scores [all pillars except evidence_maturity] + evidence_maturity*0.6) / 7.6). Step 2: verified_effectiveness = clamp(arithmetic mean of all 8 pillar scores). Step 3: apply event-level caps IN ORDER (see eventLevelCaps; apply_cap only binds when the current value strictly exceeds the cap). Step 4: ranking_score (also exposed as overall_score) = clamp(verified_effectiveness*0.65 + agenda_design_potential*0.35). There is NO Bayesian shrinkage, prior, or confidence-based score adjustment; confidence labels are informational only.

## The eight scoring pillars: definitions, weights, corpus averages
### participation_architecture (Participation)
Ratio of participatory minutes to broadcast minutes, and breadth of contribution. "People doing, not watching."
Corpus average: 29.2. Weight: 0.125 of verified_effectiveness; mechanism pillar, weight 0.1316 of agenda_design_potential.
Computed from a base of 8, plus:
  + participant_work_share (derived_count): up to 62 points
  + interactive_count (derived_count): up to 3.5 points, capped at 16
  + participant_work (source_signal): up to 5 points, capped at 10
  + format_variety (derived_count): up to 1.6 points, capped at 8
  cap 25 when: stage_only
  cap 30 when: NOT stage_only AND passive_share >= 0.8

### follow_through (Follow-through)
Named owners, dates, if-then commitments, and visible progress tracking. "Intentions become action."
Corpus average: 5.8. Weight: 0.125 of verified_effectiveness; mechanism pillar, weight 0.1316 of agenda_design_potential.
Computed from a base of 5, plus:
  + commitment (source_signal): up to 11 points, capped at 34
  + follow_up (source_signal): up to 14 points, capped at 28
  + tracking (source_signal): up to 9 points, capped at 18
  + commitment_item_count (derived_count): up to 4 points, capped at 12
  + commitment (inferred_signal): up to 2 points, capped at 8
  cap 28 when: source_signals.commitment + source_signals.follow_up == 0

### problem_specificity (Problem fit)
One costly named problem and deliberately expanded objectives. "The work has a target."
Corpus average: 30.1. Weight: 0.125 of verified_effectiveness; mechanism pillar, weight 0.1316 of agenda_design_potential.
Computed from a base of 10, plus:
  + problem (source_signal): up to 9 points, capped at 38
  + impact (source_signal): up to 5 points, capped at 16
  + format_variety (derived_count): up to 2 points, capped at 14
  + event_intent_confident (event_metadata_bonus): up to 10 points, capped at 10
  + category_known (event_metadata_bonus): up to 6 points, capped at 6
  cap 48 when: source_signals.problem == 0

### personalization (Personalization)
Content and connection tailored to attendee goals, not one-size delivery. "Usefulness beats enjoyment."
Corpus average: 23.2. Weight: 0.125 of verified_effectiveness; mechanism pillar, weight 0.1316 of agenda_design_potential.
Computed from a base of 8, plus:
  + personalization (source_signal): up to 13 points, capped at 40
  + format_variety (derived_count): up to 2.2 points, capped at 16
  + participant_types_inferred_count (event_metadata): up to 2 points, capped at 12
  + roundtable_count*3 + workshop_count*2 (derived_count_composite): up to 1 points, capped at 10
  cap 48 when: source_signals.personalization == 0

### network_design (Network design)
Deliberate bridging across groups, roles, proximity, and weak ties. "The right people connect."
Corpus average: 17.5. Weight: 0.125 of verified_effectiveness; mechanism pillar, weight 0.1316 of agenda_design_potential.
Computed from a base of 6, plus:
  + network_design (source_signal): up to 10 points, capped at 38
  + network_count (derived_count): up to 4 points, capped at 22
  + roundtable_count (derived_count): up to 5 points, capped at 14
  + generic_networking (derived_count): up to 2 points, capped at 8
  cap 45 when: source_signals.network_design == 0 AND generic_networking > 0
  cap 25 when: source_signals.network_design == 0 AND generic_networking == 0

### learning_transfer (Learning transfer)
Applied behavior at 30 to 90 days, not self-rated learning at the door. "Learning survives the room."
Corpus average: 23.1. Weight: 0.125 of verified_effectiveness; mechanism pillar, weight 0.1316 of agenda_design_potential.
Computed from a base of 7, plus:
  + transfer_count (derived_count): up to 5 points, capped at 28
  + learning_transfer + feedback (source_signal_composite): up to 10 points, capped at 28
  + follow_up (source_signal): up to 8 points, capped at 16
  + commitment (source_signal): up to 4 points, capped at 10
  + training_count (derived_count): up to 3 points, capped at 8
  cap 32 when: (source_signals.learning_transfer + source_signals.feedback) == 0 AND transfer_count == 0
  cap 35 when: stage_only

### evidence_maturity (Evidence)
Baseline, comparison, isolation, and longitudinal follow-up. "The claim can be checked."
Corpus average: 25.2. Weight: 0.125 of verified_effectiveness; weight 0.0789 of agenda_design_potential, counted at 0.6x its raw score.
Computed from a base of 0, plus:
  + source_url_present (event_metadata_bonus): up to 14 points, capped at 14
  + item_count (derived_count): up to 0.7 points, capped at 14
  + timed_count (derived_count): up to 0.8 points, capped at 10
  + baseline (source_signal): up to 12 points, capped at 22
  + tracking (source_signal): up to 11 points, capped at 22
  + impact (source_signal): up to 8 points, capped at 16
  + agenda_layer_bonus (event_metadata_bonus): up to 0 points, capped at 8
  cap 45 when: (source_signals.tracking + source_signals.baseline + source_signals.impact) == 0
  cap 35 when: source_signals.satisfaction_proxy > 0 AND (source_signals.tracking + source_signals.baseline + source_signals.impact) == 0

### future_of_work_fit (Future-fit)
Value created against time, energy, hybrid reality, and AI-augmented work. "It fits how work happens now."
Corpus average: 27.7. Weight: 0.125 of verified_effectiveness; mechanism pillar, weight 0.1316 of agenda_design_potential.
Computed from a base of 10, plus:
  + pacing_count (derived_count): up to 3 points, capped at 18
  + timed_share (derived_count): up to 22 points, capped at 16
  + future_fit (source_signal): up to 6 points, capped at 18
  + format_variety (derived_count): up to 2 points, capped at 14
  + participant_work_count (derived_count): up to 2 points, capped at 12
  cap 52 when: item_count >= 40 AND pacing_count == 0
  cap 45 when: stage_only AND source_signals.future_fit == 0

## Event-level score caps (applied in order, from the published rubric)
These caps apply after the eight pillar scores are computed, and only bind
when the current value strictly exceeds the cap (a cap can only lower a
score, never raise one). This is why an agenda can score well on individual
pillars yet still receive a low overall_score: a thin or low-quality source
page overrides the pillar math.
| Order | Applies when | Design cap | Verified cap |
|---|---|---|---|
| 1 | no_agenda (item_count == 0) | 18 | 15 |
| 2 | event.quality_rating in {'bad_page','no_agenda'} | None | 20 |
| 3 | event.quality_rating == 'lead' | None | 35 |
| 4 | stage_only | None | 32 |
| 5 | no_followup_or_tracking | None | 45 |
| 6 | each source-shape warning, in the order returned by source_shape_warnings | per warning kind | per warning kind |

## Signal detection patterns (from the published rubric)
The scorer never uses an LLM in the scoring path: every signal below is a
deterministic regex match against the agenda text (and, where noted, against
individual agenda items). This is the literal, complete, published detection
rule for every signal that feeds a pillar score.
- participant_work (feeds participation_architecture, matched against text|items): `\bhands[- ]on\b`, `\binteractive\b`, `\bworking session\b`, `\bworkshop\b`, `\blab\b`, `\bclinic\b`, `\bpractice\b`, `\bexercise\b`, `\bcase clinic\b`, `\bpeer exchange\b`, `\bco-?creat`, `\bdesign sprint\b`, `\bhackathon\b`
- commitment (feeds follow_through, matched against text|items): `\b(?:create|develop|draft|build|leave with|produce)\b.{0,50}\baction plan\b`, `\baction planning\b`, `\b(?:clear|concrete|practical|actionable|specific)\s+next steps?\b`, `\bnext steps?\s+(?:for|to)\s+(?:implement|apply|sustain|advance|strengthen|improve|continue)\b`, `\baccountability plan\b`, `\bassigned owners?\b`, `\bresponsible parties\b`, `\bby\s+(?:monday|tuesday|wednesday|thursday|friday|saturday|sunday|jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec|\d{1,2}/\d{1,2})`, `\bworking group\b`, `\bimplementation plan\b`
- follow_up (feeds follow_through, matched against text): `\bafter[- ]action\b`, `\bpost[- ]event\s+(?:evaluation|survey|follow[- ]up|session|report|resources?|action|check[- ]in)\b`, `\b(?:30|60|90)[- ]day\s+(?:plan|follow[- ]up|check[- ]in|sprint|review)\b`, `\b(?:follow[- ]up|check[- ]in)\s+(?:session|meeting|call|webinar)\b`, `\breconven(?:e|ing)\b.{0,80}\bfollow[- ]up\b`, `\bcommunity of practice\b.{0,80}\b(?:continue|ongoing|after|post[- ]event|follow[- ]up)\b`
- feedback (feeds learning_transfer, matched against text|items): `\bfeedback\b`, `\bcritique\b`, `\bcoaching\b`, `\breview session\b`, `\bassessment\b`, `\bevaluation form\b`, `\bpost[- ]event evaluation\b`, `\bsurvey\b`
- baseline (feeds evidence_maturity, matched against text): `\bbaseline\s+(?:data|measurements?|metrics?|assessment|survey|results?|capacity)\b`, `\bpre[- ]event\b`, `\bpre[- ]survey\b`, `\bpre[- ]assessment\b`, `\bbenchmark\s+(?:assessment|data|metrics?|results?|survey)\b`, `\bcontrol group\b`, `\bbefore and after\b`, `\bpre/post\b`
- tracking (feeds evidence_maturity, matched against text): `\btracked outcomes?\b`, `\bprogress tracking\b`, `\bimpact report\b`, `\boutcomes report\b`, `\bmeasured impact\b`, `\bmetrics dashboard\b`, `\blongitudinal\s+(?:follow[- ]up|tracking|evaluation|study|outcomes?|analysis)\b`, `\bresults over time\b`, `\bimpact validation\b`, `\boutcome validation\b`, `\bvalidated (?:outcomes?|impact|results?)\b`
- impact (feeds evidence_maturity, matched against text): `\bpublished (?:report|proceedings|findings|results?|papers?)\b`, `\bgrant (?:awarded|funded|secured|received)\b`, `\b(?:awarded|secured|received)\s+(?:a\s+)?grant\b`, `\bfrom pilot to scale\b`, `\bpilot(?:ed)?\b.{0,60}\b(?:results?|outcomes?|scale|implementation|launched)\b`, `\bpartnerships?\b.{0,60}\b(?:formed|launched|announced|resulted|forged|implemented)\b.{0,60}\b(?:outcomes?|results?|impact|funded|scaled|launched)\b`, `\b(?:formed|launched|announced|forged)\b.{0,40}\bpartnerships?\b.{0,60}\b(?:outcomes?|results?|impact|funded|scaled|launched)\b`, `\badoption (?:rate|metrics?|results?)\b`, `\brevenue (?:growth|generated|increase|impact)\b`, `\bROI\b.{0,40}\b(?:measured|results?|case study|analysis|evidence|impact|achieved|delivered)\b`, `\breturn on investment\b.{0,40}\b(?:measured|results?|case study|analysis|evidence|impact|achieved|delivered)\b`, `\bbehavior change\b`, `\bpolicy\b.+\bpassed\b`
- problem (feeds problem_specificity, matched against text): `\bproblem\b`, `\bchallenge\b`, `\bcostly\b`, `\bobjective\b`, `\bgoal\b`, `\boutcome\b`, `\bsolve\b`, `\bdecision criteria\b`, `\bpriorit`, `\broadmap\b`
- personalization (feeds personalization, matched against text): `\bpersonalized\b`, `\btailored\b`, `\bcurated\b`, `\brole[- ]based\b`, `\bchoose your\b`, `\bcustom agenda\b`, `\bmatchmaking\b`, `\bpre[- ]work\b`, `\battendee goals?\b`, `\blearning path\b`
- network_design (feeds network_design, matched against text|items): `\bnetworking\b`, `\bmatchmaking\b`, `\bintroductions?\b`, `\bspeed networking\b`, `\bpeer exchange\b`, `\broundtable\b`, `\bsmall group\b`, `\bmentor\b`, `\bcommunity\b`, `\bbridg`, `\bcross[- ]sector\b`, `\bcross[- ]functional\b`, `\bweak ties?\b`
- learning_transfer (feeds learning_transfer, matched against text|items): `\bapply\b`, `\bapplication\b`, `\bpractice\b`, `\btemplate\b`, `\btoolkit\b`, `\bplaybook\b`, `\bimplementation\b`, `\bworkplace\b`, `\bon the job\b`, `\bcase study\b`, `\bskill\b`, `\bcompetenc`
- future_fit (feeds future_of_work_fit, matched against text): `\bAI\b`, `\bartificial intelligence\b`, `\bgenai\b`, `\bmachine learning\b`, `\bhybrid\b`, `\bremote\b`, `\basync\b`, `\baccessib`, `\btranslation\b`, `\bsummar`, `\bworkflow\b`, `\bautomation\b`, `\btime[- ]efficient\b`
- satisfaction_proxy (feeds evidence_maturity, matched against text): `\bsatisfaction\b`, `\bNPS\b`, `\bnet promoter\b`, `\bsmile sheet\b`, `\blikert\b`, `\benjoy`, `\bloved\b`

## Reproducibility notes
Recomputing a score independently from the same agenda text should match
within 1 point. Two documented sources of that tolerance:
- Rounding: Python's round() (used inside clamp) is round-half-to-even (banker's rounding); JavaScript Math.round rounds .5 up. Scores can differ by exactly 1 when an intermediate value lands on x.5 (e.g. 24.5 -> Python 24, JS 25). For exact parity implement round-half-to-even in JS.
- Per signal type per text layer, only the FIRST pattern that matches (and survives suppression) counts, and it counts exactly once. A port that counts all matches or all patterns will overscore. Pattern order matters because suppression is checked only on the first raw match found: if pattern 1 matches in a suppressed context, the signal is skipped entirely even if pattern 2 would match cleanly elsewhere.
- The full pipeline scans up to 4 text layers (source_text, adapter_source, linked companion sources, polished agenda), so source_signals values can reach ~4 and hit the higher min() term caps. A client-side grader with one pasted agenda text has only the source_text layer, so every source_signal is 0 or 1 -- reproducing the single-layer behavior exactly (this is what the test vectors exercise).

## Corpus statistics (headline, generated 2026-07-31)
- Events scored, all records: 23,624. Average GES: 21.8.
- Over 40 GES: 4.7%. Over 50 GES: 0.2%. Over 60 GES: 0.0%.
- No visible follow-up mechanism: 92.2%. Stage-only (broadcast-only,
  no interactive format): 12.1%.
- Average pillar scores, full corpus:
- participation_architecture: 29.2
- follow_through: 5.8
- problem_specificity: 30.1
- personalization: 23.2
- network_design: 17.5
- learning_transfer: 23.1
- evidence_maturity: 25.2
- future_of_work_fit: 27.7

## Gold set (1,000) versus full corpus
Gold average GES 37.0 versus corpus 21.8. Per-pillar gap
(positive means the gold set scores higher):
- participation_architecture: gold 44.4 vs corpus 29.2 (gap 15.2)
- follow_through: gold 7.5 vs corpus 5.8 (gap 1.7)
- problem_specificity: gold 41.5 vs corpus 30.1 (gap 11.4)
- personalization: gold 37.4 vs corpus 23.2 (gap 14.2)
- network_design: gold 34.0 vs corpus 17.5 (gap 16.5)
- learning_transfer: gold 43.3 vs corpus 23.1 (gap 20.2)
- evidence_maturity: gold 41.4 vs corpus 25.2 (gap 16.2)
- future_of_work_fit: gold 54.3 vs corpus 27.7 (gap 26.6)

## Top 5 percent versus bottom half: which mechanisms actually separate them
Sample: top5 n=711, population n=14219. Top5 average GES
44.8 versus population 22.0. Per-pillar gap:
- participation_architecture: top5 61.9 vs population 30.8 (gap 31.1)
- follow_through: top5 9.5 vs population 5.9 (gap 3.6)
- problem_specificity: top5 45.2 vs population 30.1 (gap 15.1)
- personalization: top5 43.7 vs population 23.1 (gap 20.6)
- network_design: top5 41.6 vs population 18.0 (gap 23.6)
- learning_transfer: top5 56.3 vs population 23.4 (gap 32.9)
- evidence_maturity: top5 41.1 vs population 24.6 (gap 16.5)
- future_of_work_fit: top5 57.1 vs population 27.2 (gap 29.9)

Mechanism prevalence, full corpus vs top5 vs bottom50 (the gap column is
top5_share_pct minus bottom50_share_pct; the biggest gaps are the strongest
differentiators of an effective gathering):
| Mechanism | Overall | Top 5% | Bottom 50% | Gap |
|---|---|---|---|---|
| Participant work | 45.6% | 100.0% | 7.2% | 92.8 |
| Learning transfer | 70.4% | 100.0% | 44.3% | 55.7 |
| Network design | 66.3% | 99.2% | 46.2% | 53.0 |
| Feedback | 24.5% | 61.6% | 12.6% | 49.0 |
| Personalization | 16.2% | 46.0% | 8.4% | 37.6 |
| Commitments | 5.5% | 23.1% | 1.9% | 21.2 |
| Impact evidence | 1.5% | 8.4% | 0.5% | 7.9 |
| Tracking | 1.6% | 7.7% | 0.3% | 7.4 |
| Baseline | 1.5% | 2.8% | 0.4% | 2.4 |
| Follow-up | 0.3% | 1.0% | 0.1% | 0.9 |

## Score distribution (GES bands, all 23,624 records)
- 0-4: 0 events (0.0%)
- 5-9: 1,218 events (5.2%)
- 10-14: 5,278 events (22.3%)
- 15-19: 5,111 events (21.6%)
- 20-24: 3,780 events (16.0%)
- 25-29: 3,138 events (13.3%)
- 30-34: 2,314 events (9.8%)
- 35-39: 1,470 events (6.2%)
- 40-44: 768 events (3.3%)
- 45-49: 471 events (2.0%)
- 50-54: 64 events (0.3%)
- 55-59: 12 events (0.1%)
- 60-64: 0 events (0.0%)
- 65-69: 0 events (0.0%)
- 70-74: 0 events (0.0%)
- 75-79: 0 events (0.0%)
- 80-84: 0 events (0.0%)
- 85-89: 0 events (0.0%)
- 90-94: 0 events (0.0%)
- 95-100: 0 events (0.0%)

## Score distribution (GES bands, 14,219 de-duplicated representative records)
One row per distinct source page (23,624 all_records includes re-crawled
duplicates of the same source; representative_records collapses those).
- 0-4: 0 events (0.0%)
- 5-9: 819 events (5.8%)
- 10-14: 3,173 events (22.3%)
- 15-19: 2,978 events (20.9%)
- 20-24: 2,208 events (15.5%)
- 25-29: 1,665 events (11.7%)
- 30-34: 1,483 events (10.4%)
- 35-39: 991 events (7.0%)
- 40-44: 549 events (3.9%)
- 45-49: 301 events (2.1%)
- 50-54: 43 events (0.3%)
- 55-59: 9 events (0.1%)
- 60-64: 0 events (0.0%)
- 65-69: 0 events (0.0%)
- 70-74: 0 events (0.0%)
- 75-79: 0 events (0.0%)
- 80-84: 0 events (0.0%)
- 85-89: 0 events (0.0%)
- 90-94: 0 events (0.0%)
- 95-100: 0 events (0.0%)

## Corpus statistics by category (overview)
| Category | N | Avg GES | No follow-up | Top mechanisms |
|---|---|---|---|---|
| Unknown | 4,456 | 18.4 | 91.8% | Learning transfer 64.5%, Network design 54.8% |
| Technology / AI / Startup | 3,229 | 26.6 | 89.4% | Learning transfer 83.6%, Network design 74.7% |
| Academic / Research / Science | 1,172 | 24.9 | 85.7% | Learning transfer 73.5%, Network design 67.2% |
| Education / Training / Career | 1,095 | 25.2 | 88.9% | Learning transfer 80.6%, Network design 79.4% |
| Trade Show / Expo | 1,056 | 16.6 | 95.8% | Learning transfer 50.8%, Network design 49.7% |
| Culture / Media / Festival | 992 | 19.0 | 92.7% | Network design 66.0%, Learning transfer 52.2% |
| Business / Finance / Marketing | 733 | 21.0 | 90.0% | Network design 73.7%, Learning transfer 64.0% |
| Public Sector / Policy / Civic | 509 | 22.1 | 89.0% | Network design 75.2%, Learning transfer 73.5% |
| Health / Wellness / Sports | 465 | 25.2 | 85.4% | Network design 75.7%, Learning transfer 74.6% |
| Nonprofit / Social Impact | 449 | 26.2 | 83.5% | Network design 92.0%, Learning transfer 88.9% |

## Corpus statistics by category (full pillar detail)
#### Unknown (n=4,456, avg GES 18.4)
  participation_architecture: 25.6
  follow_through: 5.7
  problem_specificity: 24.1
  personalization: 20.4
  network_design: 13.6
  learning_transfer: 19.0
  evidence_maturity: 21.2
  future_of_work_fit: 21.5
  no visible follow-up: 91.8%, stage-only: 8.7%
  most common mechanisms: Learning transfer 64.5%, Network design 54.8%, Participant work 36.4%

#### Technology / AI / Startup (n=3,229, avg GES 26.6)
  participation_architecture: 36.3
  follow_through: 6.1
  problem_specificity: 34.8
  personalization: 26.5
  network_design: 23.2
  learning_transfer: 28.5
  evidence_maturity: 29.5
  future_of_work_fit: 35.2
  no visible follow-up: 89.4%, stage-only: 18.9%
  most common mechanisms: Learning transfer 83.6%, Network design 74.7%, Participant work 56.9%

#### Academic / Research / Science (n=1,172, avg GES 24.9)
  participation_architecture: 39.2
  follow_through: 6.6
  problem_specificity: 32.9
  personalization: 25.6
  network_design: 17.8
  learning_transfer: 29.5
  evidence_maturity: 26.8
  future_of_work_fit: 29.8
  no visible follow-up: 85.7%, stage-only: 6.0%
  most common mechanisms: Learning transfer 73.5%, Network design 67.2%, Participant work 57.3%

#### Education / Training / Career (n=1,095, avg GES 25.2)
  participation_architecture: 40.3
  follow_through: 6.1
  problem_specificity: 34.0
  personalization: 25.4
  network_design: 19.3
  learning_transfer: 30.5
  evidence_maturity: 25.7
  future_of_work_fit: 29.5
  no visible follow-up: 88.9%, stage-only: 10.2%
  most common mechanisms: Learning transfer 80.6%, Network design 79.4%, Participant work 62.0%

#### Trade Show / Expo (n=1,056, avg GES 16.6)
  participation_architecture: 18.4
  follow_through: 5.3
  problem_specificity: 27.8
  personalization: 18.6
  network_design: 14.5
  learning_transfer: 15.4
  evidence_maturity: 19.7
  future_of_work_fit: 22.2
  no visible follow-up: 95.8%, stage-only: 11.1%
  most common mechanisms: Learning transfer 50.8%, Network design 49.7%, Participant work 19.1%

#### Culture / Media / Festival (n=992, avg GES 19.0)
  participation_architecture: 24.8
  follow_through: 5.7
  problem_specificity: 29.7
  personalization: 20.2
  network_design: 16.0
  learning_transfer: 16.2
  evidence_maturity: 23.2
  future_of_work_fit: 24.2
  no visible follow-up: 92.7%, stage-only: 9.9%
  most common mechanisms: Network design 66.0%, Learning transfer 52.2%, Participant work 35.6%

#### Business / Finance / Marketing (n=733, avg GES 21.0)
  participation_architecture: 28.5
  follow_through: 5.7
  problem_specificity: 32.1
  personalization: 23.1
  network_design: 19.4
  learning_transfer: 20.3
  evidence_maturity: 22.1
  future_of_work_fit: 24.9
  no visible follow-up: 90.0%, stage-only: 12.6%
  most common mechanisms: Network design 73.7%, Learning transfer 64.0%, Participant work 38.3%

#### Public Sector / Policy / Civic (n=509, avg GES 22.1)
  participation_architecture: 29.7
  follow_through: 5.9
  problem_specificity: 32.8
  personalization: 21.8
  network_design: 19.2
  learning_transfer: 24.0
  evidence_maturity: 24.7
  future_of_work_fit: 26.3
  no visible follow-up: 89.0%, stage-only: 18.7%
  most common mechanisms: Network design 75.2%, Learning transfer 73.5%, Participant work 42.4%

#### Health / Wellness / Sports (n=465, avg GES 25.2)
  participation_architecture: 36.2
  follow_through: 6.8
  problem_specificity: 33.7
  personalization: 26.2
  network_design: 21.8
  learning_transfer: 27.3
  evidence_maturity: 27.5
  future_of_work_fit: 30.6
  no visible follow-up: 85.4%, stage-only: 13.1%
  most common mechanisms: Network design 75.7%, Learning transfer 74.6%, Participant work 53.1%

#### Nonprofit / Social Impact (n=449, avg GES 26.2)
  participation_architecture: 37.1
  follow_through: 6.8
  problem_specificity: 34.3
  personalization: 26.5
  network_design: 25.1
  learning_transfer: 31.4
  evidence_maturity: 29.2
  future_of_work_fit: 33.2
  no visible follow-up: 83.5%, stage-only: 5.6%
  most common mechanisms: Network design 92.0%, Learning transfer 88.9%, Participant work 77.7%

## Corpus statistics by event type (overview)
| Event type | N | Avg GES | No follow-up | Top mechanisms |
|---|---|---|---|---|
| applied learning or working session | 2,916 | 29.9 | 97.1% | Learning transfer 100.0%, Participant work 100.0% |
| mixed-format event agenda | 2,282 | 19.3 | 96.1% | Learning transfer 67.3%, Network design 63.5% |
| agenda lead | 2,115 | 11.0 | 97.4% | Network design 31.0%, Learning transfer 27.7% |
| broadcast-heavy conference | 1,542 | 24.7 | 96.8% | Learning transfer 81.5%, Network design 78.1% |
| lead or directory source | 1,346 | 19.9 | 92.9% | Network design 66.3%, Learning transfer 66.1% |
| action-oriented convening | 926 | 32.3 | 0.0% | Learning transfer 91.6%, Network design 91.4% |
| technology ai startup agenda | 647 | 19.3 | 97.7% | Learning transfer 68.3%, Network design 54.6% |
| culture media festival agenda | 444 | 16.4 | 99.3% | Network design 71.4%, Learning transfer 37.2% |
| relationship-building convening | 422 | 30.2 | 91.7% | Network design 100.0%, Learning transfer 87.0% |
| trade show expo agenda | 326 | 16.5 | 99.1% | Learning transfer 55.5%, Network design 55.5% |
| academic research science agenda | 302 | 20.0 | 97.7% | Network design 68.9%, Learning transfer 58.6% |
| education training career agenda | 269 | 21.2 | 99.3% | Network design 78.4%, Learning transfer 73.2% |
| business finance marketing agenda | 251 | 19.4 | 97.2% | Network design 77.7%, Learning transfer 60.2% |
| public sector policy civic agenda | 162 | 19.2 | 93.2% | Network design 69.1%, Learning transfer 66.0% |
| health wellness sports agenda | 116 | 19.1 | 98.3% | Network design 62.9%, Learning transfer 56.9% |
| showcase or expo program | 111 | 24.3 | 90.1% | Network design 85.6%, Learning transfer 82.0% |

## Corpus statistics by event type (full pillar detail)
#### applied learning or working session (n=2,916, avg GES 29.9)
  participation_architecture: 65.9
  follow_through: 5.0
  problem_specificity: 33.2
  personalization: 27.6
  network_design: 18.9
  learning_transfer: 36.1
  evidence_maturity: 25.7
  future_of_work_fit: 30.9
  no visible follow-up: 97.1%, stage-only: 0.0%
  most common mechanisms: Learning transfer 100.0%, Participant work 100.0%, Network design 74.9%

#### mixed-format event agenda (n=2,282, avg GES 19.3)
  participation_architecture: 20.7
  follow_through: 5.0
  problem_specificity: 27.2
  personalization: 22.0
  network_design: 15.8
  learning_transfer: 18.6
  evidence_maturity: 25.0
  future_of_work_fit: 25.9
  no visible follow-up: 96.1%, stage-only: 13.9%
  most common mechanisms: Learning transfer 67.3%, Network design 63.5%, Participant work 32.6%

#### agenda lead (n=2,115, avg GES 11.0)
  participation_architecture: 9.0
  follow_through: 5.2
  problem_specificity: 17.9
  personalization: 14.6
  network_design: 9.1
  learning_transfer: 10.8
  evidence_maturity: 14.2
  future_of_work_fit: 12.6
  no visible follow-up: 97.4%, stage-only: 0.0%
  most common mechanisms: Network design 31.0%, Learning transfer 27.7%, Feedback 10.2%

#### broadcast-heavy conference (n=1,542, avg GES 24.7)
  participation_architecture: 22.1
  follow_through: 5.0
  problem_specificity: 34.5
  personalization: 25.8
  network_design: 23.4
  learning_transfer: 26.2
  evidence_maturity: 32.9
  future_of_work_fit: 36.4
  no visible follow-up: 96.8%, stage-only: 45.6%
  most common mechanisms: Learning transfer 81.5%, Network design 78.1%, Participant work 51.1%

#### lead or directory source (n=1,346, avg GES 19.9)
  participation_architecture: 27.4
  follow_through: 5.7
  problem_specificity: 30.1
  personalization: 22.2
  network_design: 16.9
  learning_transfer: 21.1
  evidence_maturity: 22.2
  future_of_work_fit: 26.1
  no visible follow-up: 92.9%, stage-only: 12.3%
  most common mechanisms: Network design 66.3%, Learning transfer 66.1%, Participant work 36.6%

#### action-oriented convening (n=926, avg GES 32.3)
  participation_architecture: 45.8
  follow_through: 18.1
  problem_specificity: 37.2
  personalization: 30.1
  network_design: 25.3
  learning_transfer: 39.8
  evidence_maturity: 33.3
  future_of_work_fit: 37.0
  no visible follow-up: 0.0%, stage-only: 1.0%
  most common mechanisms: Learning transfer 91.6%, Network design 91.4%, Commitments 75.5%

#### technology ai startup agenda (n=647, avg GES 19.3)
  participation_architecture: 17.3
  follow_through: 5.0
  problem_specificity: 32.8
  personalization: 20.2
  network_design: 14.8
  learning_transfer: 17.3
  evidence_maturity: 23.0
  future_of_work_fit: 25.6
  no visible follow-up: 97.7%, stage-only: 17.5%
  most common mechanisms: Learning transfer 68.3%, Network design 54.6%, Participant work 19.2%

#### culture media festival agenda (n=444, avg GES 16.4)
  participation_architecture: 18.4
  follow_through: 5.0
  problem_specificity: 30.6
  personalization: 17.4
  network_design: 14.1
  learning_transfer: 12.0
  evidence_maturity: 22.4
  future_of_work_fit: 21.9
  no visible follow-up: 99.3%, stage-only: 7.0%
  most common mechanisms: Network design 71.4%, Learning transfer 37.2%, Participant work 23.0%

#### relationship-building convening (n=422, avg GES 30.2)
  participation_architecture: 31.2
  follow_through: 5.0
  problem_specificity: 36.9
  personalization: 32.1
  network_design: 45.9
  learning_transfer: 24.1
  evidence_maturity: 32.8
  future_of_work_fit: 42.3
  no visible follow-up: 91.7%, stage-only: 28.2%
  most common mechanisms: Network design 100.0%, Learning transfer 87.0%, Participant work 55.0%

#### trade show expo agenda (n=326, avg GES 16.5)
  participation_architecture: 14.9
  follow_through: 5.0
  problem_specificity: 30.4
  personalization: 18.5
  network_design: 14.5
  learning_transfer: 14.4
  evidence_maturity: 21.0
  future_of_work_fit: 24.4
  no visible follow-up: 99.1%, stage-only: 9.2%
  most common mechanisms: Learning transfer 55.5%, Network design 55.5%, Personalization 14.4%

#### academic research science agenda (n=302, avg GES 20.0)
  participation_architecture: 23.0
  follow_through: 5.0
  problem_specificity: 33.7
  personalization: 23.2
  network_design: 17.2
  learning_transfer: 19.5
  evidence_maturity: 24.0
  future_of_work_fit: 25.6
  no visible follow-up: 97.7%, stage-only: 7.3%
  most common mechanisms: Network design 68.9%, Learning transfer 58.6%, Feedback 32.5%

#### education training career agenda (n=269, avg GES 21.2)
  participation_architecture: 23.0
  follow_through: 5.0
  problem_specificity: 34.6
  personalization: 22.9
  network_design: 17.7
  learning_transfer: 22.3
  evidence_maturity: 24.4
  future_of_work_fit: 28.7
  no visible follow-up: 99.3%, stage-only: 20.8%
  most common mechanisms: Network design 78.4%, Learning transfer 73.2%, Participant work 36.1%

#### business finance marketing agenda (n=251, avg GES 19.4)
  participation_architecture: 19.4
  follow_through: 5.0
  problem_specificity: 33.8
  personalization: 22.0
  network_design: 18.6
  learning_transfer: 16.4
  evidence_maturity: 22.6
  future_of_work_fit: 26.0
  no visible follow-up: 97.2%, stage-only: 15.1%
  most common mechanisms: Network design 77.7%, Learning transfer 60.2%, Participant work 25.1%

#### public sector policy civic agenda (n=162, avg GES 19.2)
  participation_architecture: 18.7
  follow_through: 5.0
  problem_specificity: 34.1
  personalization: 20.1
  network_design: 17.2
  learning_transfer: 18.7
  evidence_maturity: 24.2
  future_of_work_fit: 25.6
  no visible follow-up: 93.2%, stage-only: 30.2%
  most common mechanisms: Network design 69.1%, Learning transfer 66.0%, Feedback 23.5%

#### health wellness sports agenda (n=116, avg GES 19.1)
  participation_architecture: 19.4
  follow_through: 5.0
  problem_specificity: 32.5
  personalization: 21.6
  network_design: 18.4
  learning_transfer: 17.1
  evidence_maturity: 25.5
  future_of_work_fit: 26.4
  no visible follow-up: 98.3%, stage-only: 20.7%
  most common mechanisms: Network design 62.9%, Learning transfer 56.9%, Feedback 23.3%

#### showcase or expo program (n=111, avg GES 24.3)
  participation_architecture: 34.6
  follow_through: 5.0
  problem_specificity: 35.5
  personalization: 27.9
  network_design: 23.3
  learning_transfer: 25.0
  evidence_maturity: 27.8
  future_of_work_fit: 28.3
  no visible follow-up: 90.1%, stage-only: 0.0%
  most common mechanisms: Network design 85.6%, Learning transfer 82.0%, Participant work 41.4%

## Sample exercise prescriptions per pillar
Real rows from the same file the get_exercises_for_pillar tool reads
(https://unitedwetransform.com/api/pillar-exercises.json), five examples per pillar so an agent can
see the actual shape before calling the tool:
- personalization (5 total): Journey Mapping • WorkshopBank; Group Energiser & Perspective-Changing Exercise - Paradigm Shift; Photo Storytelling; Assume Beginner’s Mindset; The Faces
- participation_architecture (5 total): Make A World; Awestruck 3 Minutes; Spectrum Mapping; association - Project of How; How Might We…
- problem_specificity (5 total): 0 - “Two Person Scene” Theory; Challenge Axis - Knowmium: The Un-Training Consultancy; Fist to Five • WorkshopBank; Polarity Mapping • WorkshopBank; Airplane metaphor game
- network_design (5 total): Fun Trust-Building Group Game for Team Connection - Izzat You?; Exciting Group Trust Activity for Building Confidence: Levitation; PFP: September 2005; Charismatic Voices; Online Group Energiser to Boost Connection & Creativity - Mandala
- follow_through (5 total): Mental Toughness Workshop • WorkshopBank; 15% Solutions • WorkshopBank; Network Patches; Buzby Burkley exercise; Day In The Life Of (DILO) • WorkshopBank
- learning_transfer (5 total): Team Challenge Activity for Group Problem-Solving: Don't Touch Me; Thrilling, Quick Elimination Game to Boost Energy - Match & Out; Creative Brain Teaser & Fun Geometric Team Puzzle - T-Puzzle; Fun No-Prop Energiser Game Using Imaginary Ropes - Skipping Rope; Fun Individual Challenge to Explore Creativity - Pass Through

The full exercise library (2,093 exercises, https://unitedwetransform.com/api/exercises-index.json)
by facilitation stage:
- reflection_and_learning: 960 exercises
- relationship_mapping: 623 exercises
- divergent_ideation: 246 exercises
- closure_and_transfer: 101 exercises
- context_sharing: 63 exercises
- psychological_safety: 40 exercises
- convergent_alignment: 39 exercises
- constructive_conflict: 21 exercises

## The gold leaderboard, top 100 of 1,000, with full pillar breakdown
| Rank | Title | Category | GES | Particip. | Follow-thr. | Problem | Personal. | Network | Learn.transf. | Evidence | Future-fit |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Learning Forward Annual Conference 2026 | Education / Training / Career | 59 | 57 | 42 | 63 | 49 | 32 | 79 | 65 | 72 |
| 2 | RISE National | Health / Wellness / Sports | 58 | 57 | 5 | 63 | 72 | 66 | 58 | 54 | 73 |
| 3 | Midwest Climate Collaborative 2026 | Nonprofit / Social Impact | 57 | 72 | 16 | 52 | 46 | 57 | 65 | 35 | 76 |
| 4 | 2026 Conference - Agenda at a Glance | Unknown | 57 | 81 | 28 | 48 | 46 | 38 | 71 | 60 | 66 |
| 5 | SOPHE Annual Conference | Trade Show / Expo | 56 | 72 | 33 | 53 | 40 | 62 | 79 | 62 | 67 |
| 6 | 2026 National Forum on Education Policy | Technology / AI / Startup | 54 | 61 | 16 | 49 | 66 | 66 | 62 | 35 | 79 |
| 7 | PEAK 2025 Annual Convening | Nonprofit / Social Impact | 54 | 57 | 16 | 49 | 50 | 56 | 44 | 58 | 76 |
| 8 | Caranational 2026 | Technology / AI / Startup | 54 | 68 | 16 | 47 | 50 | 54 | 54 | 35 | 68 |
| 9 | AAS Meeting | Education / Training / Career | 53 | 84 | 16 | 56 | 54 | 50 | 73 | 45 | 59 |
| 10 | Education Commission of the States | Technology / AI / Startup | 53 | 30 | 16 | 58 | 70 | 66 | 70 | 35 | 82 |
| 11 | 2026 National Forum | Technology / AI / Startup | 53 | 30 | 16 | 58 | 70 | 66 | 70 | 35 | 82 |
| 12 | CHANGE 2026 Conference 15-16 October 2026 | Technology / AI / Startup | 53 | 76 | 5 | 48 | 59 | 44 | 58 | 54 | 57 |
| 13 | 2026 Equal Justice Conference: Schedule - 2026 Equal Justice Conference - Learn more here. - 2022 | Technology / AI / Startup | 53 | 98 | 16 | 39 | 50 | 60 | 62 | 31 | 43 |
| 14 | ASCD Annual Conference 26 | Technology / AI / Startup | 52 | 52 | 16 | 59 | 50 | 16 | 64 | 66 | 52 |
| 15 | Conferences 2026 | Technology / AI / Startup | 51 | 58 | 16 | 58 | 49 | 54 | 75 | 35 | 73 |
| 16 | 2026 Retrofit Canada Conference | Energy / Industrial / Infrastructure | 51 | 57 | 16 | 43 | 59 | 59 | 67 | 35 | 82 |
| 17 | Leadership Conference 2026 - Chief Executive | Technology / AI / Startup | 51 | 68 | 5 | 47 | 50 | 57 | 55 | 58 | 80 |
| 18 | Evolution Conference Information | Education / Training / Career | 51 | 72 | 16 | 43 | 43 | 56 | 59 | 45 | 77 |
| 19 | Social Impact in the Regions 2023 | Nonprofit / Social Impact | 51 | 65 | 16 | 52 | 45 | 50 | 75 | 35 | 73 |
| 20 | Smart City Expo 2025 | Public Sector / Policy / Civic | 51 | 48 | 23 | 50 | 31 | 40 | 63 | 76 | 52 |
| 21 | IABC World Conference 2026 Home | Business / Finance / Marketing | 51 | 30 | 5 | 57 | 49 | 54 | 40 | 54 | 74 |
| 22 | Convention 2026 | Academic / Research / Science | 51 | 99 | 16 | 30 | 39 | 42 | 73 | 38 | 40 |
| 23 | Executive Platforms 2026 | Technology / AI / Startup | 50 | 55 | 5 | 48 | 55 | 56 | 55 | 54 | 80 |
| 24 | Conference 2026 | Nonprofit / Social Impact | 50 | 88 | 16 | 46 | 48 | 62 | 59 | 42 | 46 |
| 25 | Home - 2026 National Conference - Society for Corporate Governance | Technology / AI / Startup | 50 | 58 | 16 | 52 | 44 | 68 | 46 | 45 | 82 |
| 26 | Conference - Workshop - Program Staff | Nonprofit / Social Impact | 50 | 69 | 33 | 49 | 38 | 57 | 63 | 40 | 58 |
| 27 | PyCon US - Open Spaces (Friday-Sunday) | Technology / AI / Startup | 50 | 30 | 33 | 43 | 38 | 66 | 56 | 35 | 69 |
| 28 | DEF CON 33 - Workshops | Technology / AI / Startup | 50 | 91 | 16 | 46 | 46 | 16 | 62 | 35 | 59 |
| 29 | Womens Forum 2027 | Academic / Research / Science | 50 | 59 | 16 | 49 | 65 | 54 | 31 | 35 | 59 |
| 30 | 2026 Congress on Healthcare Leadership | Health / Wellness / Sports | 49 | 30 | 19 | 58 | 38 | 56 | 74 | 45 | 82 |
| 31 | MassEnergize 2026 | Nonprofit / Social Impact | 49 | 57 | 39 | 43 | 55 | 60 | 56 | 43 | 44 |
| 32 | MassEnergize 2025 | Nonprofit / Social Impact | 49 | 57 | 39 | 43 | 55 | 60 | 56 | 43 | 44 |
| 33 | 18 Sep AGIS Conference 2026 Potsdam Conference | Technology / AI / Startup | 49 | 30 | 47 | 30 | 61 | 58 | 71 | 45 | 52 |
| 34 | Nursing Conference 2026 for Healthcare Professionals | Health / Wellness / Sports | 49 | 57 | 16 | 49 | 66 | 54 | 46 | 35 | 69 |
| 35 | 2026 SOPHE Annual Conference - Browse by Track | Trade Show / Expo | 49 | 30 | 16 | 48 | 37 | 48 | 73 | 54 | 52 |
| 36 | Industrial Control Systems (ICS) Cybersecurity Conference - The ... | Technology / AI / Startup | 49 | 30 | 19 | 54 | 46 | 34 | 79 | 45 | 52 |
| 37 | 2025 Ecosystem Partners | Trade Show / Expo | 49 | 71 | 16 | 45 | 44 | 60 | 57 | 31 | 38 |
| 38 | 2025 Highlights - Ecosystem Partners | Trade Show / Expo | 49 | 73 | 16 | 45 | 44 | 56 | 57 | 31 | 38 |
| 39 | Unbound Book Festival 2026 | Culture / Media / Festival | 49 | 82 | 16 | 47 | 52 | 6 | 67 | 35 | 55 |
| 40 | The AI Conference 2026 San Francisco AI Conference | Technology / AI / Startup | 49 | 66 | 5 | 49 | 59 | 56 | 58 | 45 | 81 |
| 41 | 13 Jul 2026 Badge Summit Conference | Technology / AI / Startup | 48 | 47 | 5 | 62 | 38 | 48 | 63 | 62 | 80 |
| 42 | 2026 SOPHE Annual Conference | Trade Show / Expo | 48 | 50 | 16 | 48 | 40 | 48 | 73 | 54 | 67 |
| 43 | HMP Global Events 2027 | Technology / AI / Startup | 48 | 69 | 5 | 53 | 48 | 28 | 66 | 62 | 73 |
| 44 | Devlearn 2026 | Education / Training / Career | 48 | 66 | 5 | 63 | 65 | 16 | 71 | 54 | 52 |
| 45 | Educating Leaders 2026 - AACOM | Health / Wellness / Sports | 48 | 59 | 14 | 34 | 42 | 34 | 69 | 57 | 74 |
| 46 | Asplos Conference 2025 | Education / Training / Career | 48 | 52 | 33 | 43 | 46 | 46 | 61 | 45 | 71 |
| 47 | AIIM Global Summit 2026 | Unknown | 48 | 52 | 16 | 43 | 57 | 42 | 62 | 35 | 76 |
| 48 | Conference - Workshop - 2 | Nonprofit / Social Impact | 48 | 78 | 33 | 41 | 35 | 35 | 71 | 41 | 48 |
| 49 | 2026 Marketing and Communications Summit | Business / Finance / Marketing | 48 | 30 | 5 | 54 | 36 | 53 | 45 | 54 | 71 |
| 50 | OLC Innovate 2026 - Wednesday 3/04 | Education / Training / Career | 48 | 30 | 5 | 49 | 57 | 34 | 58 | 62 | 54 |
| 51 | ARPA-H Events | Health / Wellness / Sports | 48 | 45 | 5 | 58 | 57 | 47 | 71 | 45 | 64 |
| 52 | Women's Leadership Summit 2026 - The Room - 2026 Events | Business / Finance / Marketing | 48 | 66 | 5 | 49 | 42 | 57 | 48 | 45 | 82 |
| 53 | 2026 Conference - Workshops at ICLR 2026 | Technology / AI / Startup | 47 | 77 | 5 | 63 | 45 | 16 | 63 | 54 | 65 |
| 54 | CFP 10th International Conference on Urban e-Planning, 8 - AESOP - PhD Workshop Poznan 2023 | Education / Training / Career | 47 | 70 | 5 | 45 | 39 | 60 | 45 | 54 | 65 |
| 55 | Women's Leadership Summit | Education / Training / Career | 47 | 30 | 5 | 63 | 66 | 54 | 37 | 54 | 64 |
| 56 | Euromlsys 2025 | Technology / AI / Startup | 47 | 69 | 16 | 47 | 48 | 16 | 67 | 57 | 69 |
| 57 | Explore the Full Program | Unknown | 47 | 48 | 5 | 48 | 45 | 33 | 63 | 49 | 55 |
| 58 | Learning Technologies - 2026 Programme | Technology / AI / Startup | 47 | 30 | 5 | 64 | 23 | 30 | 71 | 62 | 52 |
| 59 | Submissions 2025 | Technology / AI / Startup | 47 | 30 | 22 | 52 | 43 | 26 | 59 | 45 | 67 |
| 60 | 2026 National Conference - GEO | Nonprofit / Social Impact | 47 | 30 | 16 | 43 | 38 | 56 | 49 | 45 | 68 |
| 61 | Programming & Events ATX TV Festival | Culture / Media / Festival | 47 | 51 | 16 | 40 | 41 | 53 | 59 | 40 | 40 |
| 62 | Nonprofit Technology Conference | Technology / AI / Startup | 47 | 30 | 16 | 56 | 33 | 48 | 39 | 35 | 72 |
| 63 | HR Technology Conference | Technology / AI / Startup | 47 | 49 | 5 | 58 | 53 | 64 | 58 | 35 | 76 |
| 64 | 08 Jul FreedomFest 2026 Conference | Education / Training / Career | 47 | 56 | 5 | 52 | 42 | 62 | 45 | 35 | 70 |
| 65 | 25th Global Summit on Nursing Education and Practice (GSNEP 2026) | Health / Wellness / Sports | 47 | 71 | 5 | 40 | 72 | 50 | 69 | 35 | 66 |
| 66 | 2026 Conference on Philanthropy - Classes & Workshops - fun and energizing activities | Business / Finance / Marketing | 47 | 98 | 5 | 50 | 29 | 34 | 66 | 35 | 40 |
| 67 | Researchersfusion 2026 | Academic / Research / Science | 47 | 63 | 5 | 40 | 46 | 48 | 38 | 35 | 71 |
| 68 | Eurocities Annual Conference | Technology / AI / Startup | 47 | 54 | 5 | 49 | 49 | 50 | 58 | 45 | 74 |
| 69 | Gitexeurope 2026 | Technology / AI / Startup | 47 | 80 | 5 | 49 | 53 | 66 | 53 | 38 | 58 |
| 70 | Sydney Corporate 2026 Women Unlimited Leadership Summit | Technology / AI / Startup | 47 | 27 | 5 | 49 | 62 | 50 | 45 | 45 | 66 |
| 71 | 2026 Equal Justice Conference: Schedule - 2026 Equal Justice Conference - Learn more here. - 2021 | Technology / AI / Startup | 47 | 100 | 5 | 37 | 35 | 50 | 66 | 30 | 36 |
| 72 | Techexevent 2026 | Technology / AI / Startup | 47 | 73 | 5 | 43 | 27 | 52 | 50 | 42 | 57 |
| 73 | Submit by July 15 | Unknown | 47 | 77 | 5 | 43 | 57 | 54 | 61 | 41 | 72 |
| 74 | NC LIVE 2026 Annual Conference: Event Locked - Sched - NC LIVE 2026 Annual Conference - In-Person Workshops | Technology / AI / Startup | 47 | 78 | 5 | 47 | 61 | 42 | 69 | 35 | 66 |
| 75 | Cmu Safari 2026 | Academic / Research / Science | 47 | 77 | 5 | 52 | 65 | 38 | 63 | 43 | 70 |
| 76 | NPC26 Online | Technology / AI / Startup | 47 | 30 | 5 | 52 | 53 | 52 | 55 | 45 | 79 |
| 77 | Smart City Expo 2030 | Public Sector / Policy / Civic | 47 | 49 | 5 | 58 | 36 | 64 | 63 | 45 | 52 |
| 78 | 19 Jul 2026 CDS Annual Conference Conference | Nonprofit / Social Impact | 47 | 56 | 5 | 52 | 43 | 52 | 63 | 45 | 73 |
| 79 | 2026 NAEA National Convention: Schedule | Health / Wellness / Sports | 47 | 57 | 5 | 38 | 41 | 64 | 66 | 45 | 52 |
| 80 | Go to Europe event - Go to US event - Explore the agenda → | Unknown | 46 | 30 | 25 | 43 | 59 | 60 | 36 | 52 | 76 |
| 81 | VLDB - Overall Program Structure | Academic / Research / Science | 46 | 69 | 33 | 52 | 38 | 36 | 46 | 45 | 63 |
| 82 | AAS Meeting - AAS Meeting Sessions & Content | Education / Training / Career | 46 | 94 | 16 | 30 | 52 | 34 | 67 | 37 | 45 |
| 83 | Voices of Women 2026 - EventRaptor | Culture / Media / Festival | 46 | 62 | 16 | 49 | 55 | 36 | 73 | 35 | 53 |
| 84 | Interactive Agenda | Unknown | 46 | 57 | 16 | 41 | 35 | 56 | 59 | 41 | 66 |
| 85 | Online Learning Consortium 2026 | Education / Training / Career | 46 | 76 | 16 | 45 | 46 | 40 | 67 | 37 | 39 |
| 86 | Robotics Workshop 2026 | Academic / Research / Science | 46 | 66 | 33 | 43 | 41 | 30 | 63 | 39 | 57 |
| 87 | 2026 Schedule | Unknown | 46 | 47 | 22 | 43 | 44 | 52 | 57 | 35 | 63 |
| 88 | SIGGRAPH 2025 Conference Schedule | Unknown | 46 | 55 | 16 | 43 | 40 | 43 | 57 | 41 | 63 |
| 89 | NCSL Legislative Summit | Public Sector / Policy / Civic | 46 | 25 | 5 | 57 | 49 | 48 | 32 | 54 | 66 |
| 90 | Lifestyle Medicine Conference | Academic / Research / Science | 46 | 59 | 5 | 40 | 42 | 52 | 63 | 45 | 82 |
| 91 | InfoSec World 2026 | Health / Wellness / Sports | 46 | 61 | 5 | 43 | 44 | 61 | 69 | 42 | 68 |
| 92 | MGMA Leaders Conference | Trade Show / Expo | 46 | 57 | 5 | 34 | 44 | 64 | 63 | 45 | 82 |
| 93 | Milkeninstitute 2025 | Business / Finance / Marketing | 46 | 47 | 5 | 52 | 64 | 66 | 42 | 35 | 78 |
| 94 | FCM Annual Conference - Workshops | Technology / AI / Startup | 46 | 65 | 5 | 56 | 63 | 46 | 63 | 43 | 51 |
| 95 | ICML - WCA4Z Platform - Accelerating Legacy Code Modernization Using AI Agents, Deep Program Analysis, and - 2026 - Workshops | Academic / Research / Science | 46 | 70 | 5 | 54 | 52 | 38 | 69 | 45 | 52 |
| 96 | View Education Sessions from 2025 | Education / Training / Career | 46 | 45 | 5 | 49 | 38 | 43 | 55 | 41 | 67 |
| 97 | Wce.iafor.org | Unknown | 46 | 54 | 5 | 43 | 59 | 54 | 63 | 41 | 64 |
| 98 | Presentation Schedule | Unknown | 46 | 54 | 5 | 43 | 59 | 54 | 63 | 41 | 64 |
| 99 | Lifestyle Medicine Conference 2026 Orlando, FL and Online - Explore Workshops Now | Academic / Research / Science | 46 | 61 | 5 | 34 | 66 | 50 | 63 | 45 | 82 |
| 100 | 18 Aug Outdoor Writers Association of America (OWAA) 2026 Conference Conference | Culture / Media / Festival | 46 | 70 | 5 | 34 | 42 | 48 | 55 | 45 | 82 |
Full 1,000-row leaderboard: https://unitedwetransform.com/api/leaderboard.json. Full search facets
(same 1,000 rows): https://unitedwetransform.com/api/search-index.json. Full per-event scorecard,
including per-pillar evidence_provenance and prescriptions: https://unitedwetransform.com/api/score/{event-slug}.json.

## The 10 MCP tool contracts
POST https://unitedwetransform.com/mcp: JSON-RPC 2.0 over HTTP, streamable-HTTP transport, no auth,
no session, stateless (every request is independent). All tools carry
readOnlyHint:true. Discovery: https://unitedwetransform.com/.well-known/mcp.json and
https://unitedwetransform.com/.well-known/mcp/server-card.json. Per-client setup: https://unitedwetransform.com/use-with-ai/

### grade_agenda
Score a pasted event agenda with the deterministic UWT grading rubric. The text is processed transiently: it is never stored, never logged beyond aggregate counts, and never used to add or alter any entry in the atlas.

input_schema:
{
  "type": "object",
  "properties": {
    "agenda_text": {
      "type": "string",
      "description": "The full event agenda, one session per line. Times are optional. Example: '9:00 Opening / 10:00 Workshop: map the problem / 3:00 Commitments: owners and dates'."
    }
  },
  "required": [
    "agenda_text"
  ],
  "additionalProperties": false
}

Example call: {"agenda_text": "9:00am Opening Remarks\n10:30am Workshop..."}
Real result (this exact input, scored just now):
overall_score=30, pillars={"participation_architecture": 40, "follow_through": 19, "problem_specificity": 29, "personalization": 21, "network_design": 22, "learning_transfer": 20, "evidence_maturity": 30, "future_of_work_fit": 38}, missing_mechanisms=[]

### get_event_score
Fetch the cached scorecard for one gathering by slug from the public atlas.

input_schema:
{
  "type": "object",
  "properties": {
    "slug": {
      "type": "string",
      "description": "The event slug, i.e. the last path segment of an /events/<slug>/ URL. On a miss the response documents how to grade it yourself."
    }
  },
  "required": [
    "slug"
  ],
  "additionalProperties": false
}

Example call: {"slug": "evt_2026_learning_forward_annual_conference_2026_lfp_learningforward_org"}
Real result: {"id": "evt_2026_learning_forward_annual_conference_2026_lfp_learningforward_org", "ges": 59, "verification_tier": "gold_verified", "pillars": [{"id": "evidence_maturity", "label": "Evidence Maturity", "score": 65, "confidence": "medium", "evidence_provenance": {"source_backed": 2, "inferred": 0}, "rationale": "Source link present; 0 baseline, 1 tracking, 1 impact signals."}, {"id": "follow_through", "label": "Follow Through", "score": 42, "confidence": "medium", "evidence_provenance": {"source_backed": 2, "inferred": 1}, "rationale": "2 source-backed commitment signals, 0 follow-up signals, 1 tracking signals."}, {"id": "future_of_work_fit", "label": "Future-of-Work Fit", "score": 72, "confidence": "medium", "evidence_provenance": {"source_backed": 2, "inferred": 3}, "rationale": "2 future-of-work signals, timed share 56%, 4 pacing blocks."}, {"id": "learning_transfer", "label": "Learning Transfer", "score": 79, "confidence": "high", "evidence_provenance": {"source_backed": 4, "inferred": 8}, "rationale": "10 transfer-capable blocks, 4 source-backed transfer/feedback signals."}, {"id": "network_design", "label": "Network Design", "score": 32, "confidence": "medium", "evidence_provenance": {"source_backed": 1, "inferred": 4}, "rationale": "1 source-backed network-design signals, 4 network-capable blocks, 0 generic networking blocks."}, {"id": "participation_architecture", "label": "Participation Architecture", "score": 57, "confidence": "medium", "evidence_provenance": {"source_backed": 2, "inferred": 8}, "rationale": "9 participant-work blocks, 11 interactive blocks, passive share 70%."}, {"id": "personalization", "label": "Personalization", "score": 49, "confidence": "medium", "evidence_provenance": {"source_backed": 1, "inferred": 0}, "rationale": "1 tailoring signals and 5 inferred participant segments."}, {"id": "problem_specificity", "label": "Problem Specificity", "score": 63, "confidence": "medium", "evidence_provenance": {"source_backed": 2, "inferred": 0}, "rationale": "2 source-backed problem/objective signals; event intent confidence medium."}], "correction_url": "https://unitedwetransform.com/methodology-trust-boundary/"}

### compare_events
Compare two cached scorecards pillar by pillar, with a one-line summary of which event scores higher overall and by how much. Answers 'compare two events I am considering sponsoring' in one call.

input_schema:
{
  "type": "object",
  "properties": {
    "slug_a": {
      "type": "string",
      "description": "First event slug to compare."
    },
    "slug_b": {
      "type": "string",
      "description": "Second event slug to compare."
    }
  },
  "required": [
    "slug_a",
    "slug_b"
  ],
  "additionalProperties": false
}



### search_gatherings
Search the gold-verified gathering index by free-text query, category, and minimum GES.

input_schema:
{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "description": "Free-text match against event title, category, and type. Every word in the query must appear somewhere in those fields."
    },
    "category": {
      "type": "string",
      "description": "Restrict to one gathering category (substring match, e.g. \"Education\")."
    },
    "min_ges": {
      "type": "number",
      "minimum": 0,
      "maximum": 100,
      "description": "Minimum Gathering Effectiveness Score (0-100)."
    },
    "max_results": {
      "type": "integer",
      "minimum": 1,
      "maximum": 20,
      "default": 20,
      "description": "Maximum rows to return (capped at 20)."
    }
  },
  "additionalProperties": false
}

Example call: {"query": "climate summit", "min_ges": 40}
Returns rows shaped exactly like the leaderboard table above, filtered by tokenized title/category/event_type match and minimum GES, sorted by GES descending, each enriched with canonical and score_json.

### search
Federated search across gold-verified events, facilitation exercises, and core atlas pages, for clients (including ChatGPT deep research) that expect a generic search tool rather than search_gatherings' event-only filters. Returns citation-ready absolute URLs.

input_schema:
{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "description": "Free-text query. Every word must appear in the matched title, category, or description."
    }
  },
  "required": [
    "query"
  ],
  "additionalProperties": false
}



### fetch
Fetch the full content for one id returned by search: the complete scorecard text for an event, the summary for an exercise, or the summary for a core page. Required by ChatGPT deep research.

input_schema:
{
  "type": "object",
  "properties": {
    "id": {
      "type": "string",
      "description": "An id returned by the search tool (event:<slug>, exercise:<slug>, or page:<path>)."
    }
  },
  "required": [
    "id"
  ],
  "additionalProperties": false
}



### get_statistics
Return corpus-wide summary statistics for the atlas.

input_schema:
{
  "type": "object",
  "properties": {},
  "additionalProperties": false
}

Example call: {}
Real result: the exact object summarized in Corpus statistics above.

### get_leaderboard
Return the gold-verified leaderboard, ranked by GES, up to a requested number of rows.

input_schema:
{
  "type": "object",
  "properties": {
    "top_n": {
      "type": "integer",
      "minimum": 1,
      "maximum": 100,
      "default": 100,
      "description": "Number of top-ranked gold events to return (capped at 100)."
    }
  },
  "additionalProperties": false
}

Example call: {"top_n": 3}
Real result (first 3 of the 100 rows above): [{"rank": 1, "title": "Learning Forward Annual Conference 2026", "ges": 59}, {"rank": 2, "title": "RISE National", "ges": 58}, {"rank": 3, "title": "Midwest Climate Collaborative 2026", "ges": 57}]

### get_exercises_for_pillar
Return the suggested improvement exercises for one scoring pillar (for example follow_through or network_design).

input_schema:
{
  "type": "object",
  "properties": {
    "pillar": {
      "type": "string",
      "enum": [
        "personalization",
        "participation_architecture",
        "problem_specificity",
        "network_design",
        "follow_through",
        "learning_transfer"
      ],
      "description": "One scoring pillar that carries exercise prescriptions."
    }
  },
  "required": [
    "pillar"
  ],
  "additionalProperties": false
}

Example call: {"pillar": "personalization"}
Real result: [{"slug": "journey-mapping-workshopbank", "title": "Journey Mapping • WorkshopBank"}, {"slug": "group-energiser-perspective-changing-exercise-paradigm-shift", "title": "Group Energiser & Perspective-Changing Exercise - Paradigm Shift"}]

### make_teams
Form balanced teams or breakout rounds from a pasted participant roster (CSV or TSV with a header row). Runs the same deterministic team formation engine that ships client-side at https://unitedwetransform.com/team-creator/. Supports sliders for expertise_domain, organization, geography, seniority, gender, network_familiarity, ask_offer_fit, helping_potential, or a custom:<column> dimension, plus must_together and must_apart name constraints and multi-round breakout planning. Processed transiently: the participant list is never stored, never logged beyond an aggregate count, and never used to add or alter any entry in the atlas.

input_schema:
{
  "type": "object",
  "properties": {
    "participants_csv": {
      "type": "string",
      "description": "CSV or TSV text with a header row and at least 4 named data rows; needs a recognizable name column."
    },
    "sliders": {
      "type": "object",
      "description": "Map of dimension to an integer from -5 (group similar) to 5 (spread diverse). Dimensions: expertise_domain, organization, geography, seniority, gender, network_familiarity, ask_offer_fit, helping_potential, or a custom:<column> key naming any extra column in the CSV.",
      "additionalProperties": {
        "type": "integer",
        "minimum": -5,
        "maximum": 5
      }
    },
    "must_together": {
      "type": "array",
      "description": "Pairs of participant names to keep on the same team.",
      "items": {
        "type": "array",
        "items": {
          "type": "string"
        },
        "minItems": 2,
        "maxItems": 2
      }
    },
    "must_apart": {
      "type": "array",
      "description": "Pairs of participant names to keep on different teams.",
      "items": {
        "type": "array",
        "items": {
          "type": "string"
        },
        "minItems": 2,
        "maxItems": 2
      }
    },
    "team_size": {
      "type": "integer",
      "minimum": 2,
      "maximum": 15,
      "description": "Target people per team. Set this or num_teams."
    },
    "num_teams": {
      "type": "integer",
      "minimum": 1,
      "description": "Number of teams to form. Set this or team_size."
    },
    "rounds": {
      "type": "integer",
      "minimum": 1,
      "maximum": 8,
      "default": 1,
      "description": "Number of breakout rounds to plan, with repeat-pair avoidance across rounds."
    },
    "seed": {
      "type": "integer",
      "description": "Optional seed for reproducible team formation."
    }
  },
  "required": [
    "participants_csv"
  ],
  "additionalProperties": false
}

Example call: {"participants_csv": "name,organization,title\nAlice Anders,Acme Corp,Engineer\n...", "team_size": 4, "sliders": {"organization": -5}, "seed": 42}
Real result, round 1 teams (this exact 8-person roster, run just now): [["Alice Anders", "Bob Baker", "Dave Diaz", "Erin Evans"], ["Carol Chen", "Frank Ford", "Grace Gomez", "Henry Hall"]]

## The GET API reference
The read API is a static cache: every path below is GET-only, and scoring
itself always happens client-side or inside grade_agenda, never as a side
effect of a GET. Full machine-readable spec: https://unitedwetransform.com/api/openapi.json
- GET /api/score/{eventSlug}.json: Cached scorecard for one event (always GET this before re-deriving)
- GET /api/leaderboard.json: The 1,000 hand-verified gold events, full spread, ranked by GES
- GET /api/statistics.json: Citable aggregate statistics for the 23,624-agenda corpus
- GET /api/search-index.json: Gold-only search facets (id, title, category, ges, page)
- GET /api/exercises-index.json: 2,093 published exercises with stage, pillar, and url
- GET /api/pillar-exercises.json: Top exercise prescriptions per GES pillar
- GET /api/openapi.json: This document

## Typical agent workflows
- Grading a new agenda: call grade_agenda with the full text, one session
  per line. Read overall_score against the corpus average (21.8) and the
  score distribution above to judge where it lands, then call
  get_exercises_for_pillar for its weakest pillar to suggest a concrete fix.
- Benchmarking a known event: call search_gatherings with a few title words
  or min_ges, then get_event_score on the matched id for the full scorecard
  (pillars, evidence_provenance, prescriptions). Compare its pillar numbers
  against the category or event-type detail tables above.
- Forming workshop teams: paste an attendee roster (CSV or TSV with a header
  row) into make_teams with a team_size or num_teams and any similar-versus-
  diverse sliders; add must_together or must_apart pairs for known
  constraints. The same engine runs client-side at https://unitedwetransform.com/team-creator/.
- Citing a statistic: call get_statistics or get_leaderboard rather than
  re-deriving a number from raw event pages; every figure returned there
  matches this file's Corpus statistics section exactly, since both are
  generated from the same underlying data at build time.

## What GES is not
GES measures what a gathering disclosed and structured in its public agenda.
It does not measure event quality, attendee experience, or organizer
competence. A missing signal in a published agenda is not proof the event
lacked it. Organizers can request corrections or removal:
corrections@unitedwetransform.com.

## Emerging signals (tracked, not yet scored)
Post-event outcome evidence (searched, classified organizer_claim vs
third_party vs measured_outcome), Wayback-archived source snapshots, and
identity verification for people profiles (OpenAlex, Semantic Scholar,
Wikidata) are collected and displayed but do not move scores until the
methodology promotes them, with a public changelog. Bonus signals never
penalize non-adopters.

## Citation and correction contract
Cite "United We Transform, Gathering Effectiveness dataset (2026-07-31)" and
link the scorecard or statistic used. The dataset is CC-BY 4.0: attribution
required, reuse welcome, including AI training. Organizers and agents can
flag errors: corrections@unitedwetransform.com with the event slug, or the
correction link on any scorecard. Five business day review window. No
upsell fields exist anywhere in this surface, ever: not in scorecards, not in
tool responses, not in this file.