Add timing_window structured field to genesis.js PROPOSER_SYSTEM output schema

council rejected PROMPT reversible: simple 3h proposed 19 May 2026

What is the proposed change?

In the genesis.js PROPOSER_SYSTEM JSON output schema, add required field `timing_window` as a structured object. Schema: `{"window_state": "opening" | "open" | "stable" | "closing", "triggering_event": "<25 words: what specifically changed in the last 12 months to make this solvable or urgent now — name the technology, regulation, or market shift>", "decay_horizon": "3_months" | "12_months" | "3_years" | "structural"}`. Instruction: 'window_state=opening means a specific recent change created this gap; open means the window is present and holding; closing means competition is moving; structural means the problem has always existed and no timing advantage applies. triggering_event must be a named change — "AI agents becoming real" is too vague; "OpenAI Codex Apps GA February 2026 enabling async agent workflows" is specific. decay_horizon estimates how long before the window is saturated by incumbents or model improvements.'

Target files

hypothesis_engine/moves/genesis.js

Expected effect

Genesis outputs for commodity or evergreen problems (e.g. taxonomy normalizers, knowledge-base tools) will produce window_state=stable or structural with decay_horizon=3_months, signaling low timing defensibility. Proposals tied to genuine substrate shifts (agent-era trust gaps, EU AI Act August 2026, MCP ecosystem maturing) will produce window_state=opening with specific triggering_events. Filter_score and council can use this field to weight time-sensitive candidates higher. The field recovers the agentic_decay_forecast concept dropped from v1 without the full forecasting architecture.

Falsifier — what would prove this wrong?

Review 10 genesis outputs post-change. ≥8 must contain a triggering_event naming a specific product launch, regulation, or market event (not a generic AI trend). Back-apply to 5 recently graduated candidates: the 3 that council killed for 'no observed buyer / episodic tension / wrong timing' should produce window_state=closing or stable; the 2 ROBUST S157 candidates (ec4507 and 3656a0) should produce window_state=open or opening. If correlation between window_state and council verdict direction is below 60%, the field is not discriminating on the right dimension.

Evidence that triggered the proposal

Corpus D, red_team_reviews/s118-v0-vs-v1-gpt-5.4-pro.md: 'What did v0 lose that v1 had that actually mattered? — Explicit timing/window analysis. v1's agentic_decay_forecast idea was genuinely useful. AE does not just need good opportunities; it needs opportunities whose window is open now and not already collapsing.'
Corpus D, red_team_reviews/s118-v0-vs-v1-gemini-3.1-pro.md: 'What did v0 correctly strip from v1? (among kept items) What it lost: the timing decay forecast. Add a Window of Opportunity section to recover it.'
Corpus D, S122_TRANSCRIPT_ANALYSIS.md: 'Substrate-shift reframe… The shift is asymmetric. It opens a structurally different role for AE.' — timing of the substrate shift determines whether verifier-layer positioning has a 3-month or 3-year window; genesis doesn't capture this signal

Proposer self-score

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

Axis	Score
specificity	3
falsifier	2
solo feasible	3
blast radius	3
composability	3
reversibility	3

Disposition

Rejected at the council verdict. The two-judge council did not find the case strong enough to advance to Commander review.

Evaluation history

When	Move
2026-05-23 04:32	meta_council_verdict
2026-05-23 04:18	meta_argument
2026-05-19 12:36	red_team_kill
2026-05-19 10:54	steelman
2026-05-19 10:06	meta_filter_score
2026-05-19 09:53	meta_genesis