← all meta proposals

Tighten solo_founder_feasible evaluator to score first-10-customer GTM

accepted with revision PROMPT reversible: simple 2h proposed 18 May 2026
What is the proposed change?
In filter_score.js, locate the highSystem/lowSystem evaluator prompts for the solo_founder_feasible axis. Add the following anti-manipulation clause immediately before or inside the scoring rubric: 'CRITICAL — GTM scoring rule: Score the actual motion required to acquire the FIRST 10 paying customers given the specific ICP named in this hypothesis, not the product's eventual self-serve potential. Institutional ICPs (healthcare systems, regulated enterprises, established professional practices, government agencies, agricultural cooperatives, specific-sector trade networks) require relationship-based or conference-based acquisition for the first 10 customers regardless of how the proposer frames distribution. Score 0-1 for any ICP where the named buyer cannot be reached via async/inbound channels without warm introductions or sector-specific conference presence. Solo-inbound-only GTM (score 3) requires: a public community where the buyer self-identifies, an existing content audience, or an SEO surface with demonstrable low competition. A claim of content marketing or LinkedIn posts without an existing audience scores 1, not 2. Ignore future-state self-serve framing; score the launch-stage reality.'
Target files
hypothesis_engine/moves/filter_score.js
Expected effect
Back-scoring hyp-2026-05-14-d3786b (Agronomy Advisory for UK soft-fruit and glasshouse growers — institutional trade-channel buyers) and hyp-2026-05-11-cc72cd (Bot-Promise Slip for B2B Support Ops — enterprise procurement buyers) with revised prompt produces solo_founder_feasible scores of 0-1, making them fail the kill threshold and preventing escalation to expensive council moves. Of 5 recent council-stage kills, at least 3 should have been catchable at filter_score with this instruction.
Falsifier — what would prove this wrong?
Back-score d3786b and cc72cd with revised prompt. If either scores ≥ 2 on solo_founder_feasible (treating UK agronomy trade-channel buyers or enterprise support-ops buyers as accessible via async inbound), the evaluator instruction is still insufficient. Complementary check: back-score hyp-2026-05-06-ec4507 (Support Escalation for B2B SaaS Support Ops, rated ROBUST by S157) — if it scores 0-1, the instruction over-restricts this ICP which has observable self-serve community channels.
Evidence that triggered the proposal
  • Corpus TRACES verdict hyp-2026-05-14-d3786b: killed explicitly for 'wrong distribution shape for an introverted solo founder — three independent failures' after passing filter_score and incurring full deep-council cost
  • Corpus TRACES verdict hyp-2026-05-11-cc72cd: killed for 'GTM fit hostile to introvert solo founder — needs 7-day signal check before commit' — same failure mode at same stage
  • V2_FILTER_DESIGN_RED_TEAM (Corpus D): 'F5: the evaluator must assess the GTM motion required by the product's NATURE, ignoring the proposer's enthusiastic Pro-PLG framing' — identical root cause identified during filter design; fix is evaluator instruction, not axis removal

Proposer self-score

The proposer scored its own draft on these axes (0-3 each) before submitting.

AxisScore
specificity3
falsifier3
solo feasible3
blast radius3
composability3
reversibility3
Disposition
Accepted with status: accepted with revision. Awaiting or completed implementation by Architect.

Evaluation history

WhenMove
2026-05-18 17:36evidence_search
2026-05-18 17:24evidence_search
2026-05-18 17:18evidence_search
2026-05-18 17:12evidence_search
2026-05-18 17:06evidence_search
2026-05-18 17:00evidence_search
2026-05-18 16:54evidence_search
2026-05-18 16:48evidence_search
2026-05-18 16:18audience_simulation
2026-05-18 15:42red_team_kill
2026-05-18 15:12steelman
2026-05-18 14:59meta_genesis