August 13, 2026

Generated lab notebook from daily_summary, plan_journal, plan_delivery_log, and setpoint audit data. It is intentionally chronological and may include in-progress cycles before validation.

Planner Execution Ledger

Planner checkpoints can acknowledge that the active plan is still suitable without writing a new public plan ID. Those no-change decisions are part of the audit trail.

🌅 Morning Cycle (12:20 AM) — iris-20260813-0017

Reflection

Validating previous cycle: iris-20260812-0610

Previous hypothesis: SUNRISE plan for Wednesday 2026-08-12. Yesterday closed at score 71.4 versus a 66.3 seven-day average, with no resource score. VPD was the bottleneck: 32.9% legacy axis compliance versus 46.6% temperature compliance; VPD-high stress dominated at 14.69 h versus 10.89 h heat, with 0 h cold and VPD-low stress. Water was 325 gal versus a 239 gal seven-day average, including 282 gal of misting. This was a cooling/misting day, but electric, gas, and total-cost fields are missing, so exact utility attribution is unavailable. The midnight plan recovered VPD from about 1.21 kPa to 0.75-0.78 kPa by sunrise with zero dew-risk hours. I scored it 6/10 versus deterministic anchor 3: my premium credits measured recovery and a 9.5F minimum dew margin, while the anchor correctly penalizes 116 guardrail events, 106 held, and incomplete proof of clean daytime materialization… Result: Plan governed approximately 00:20-06:14 MDT. Window score/compliance was 57.1, below the recent 70.4 daily baseline; legacy temperature compliance was 50.9% and VPD compliance 54.6%. Heat stress was the largest window component at 1.90 h, followed by 1.43 h VPD-high, 0.20 h VPD-low, and 0 h cold. The humid forecast direction did arrive by dawn: house VPD declined from about 0.87 kPa at midnight to 0.74 kPa by 06:00 while RH rose to roughly 73%, without cold stress. Dew safety was preserved; the day later closed with 7.4F minimum dew margin and 0 dew-risk hours. However, the plan did not deliver strong corridor performance, and the predicted humidity transition was too slow to justify early suppression while the house began above the active VPD high edge. The durable operational conclusion remains to retain band-coupled moisture authority until measured VPD is inside the corridor and the VENTILATE moisture-gap alert clears, then taper for humid-night risk. Resource and reliable cost fields were unavailable, so no cost-efficiency claim is made. Score: 4/10

Hypothesis

Testing: Use a short, dew-safe band-coupled recovery window before the humid pre-dawn taper; test whether conditioning the taper on observed recovery, rather than forecast humidity alone, clears the live VPD-high gap without causing pre-dawn VPD-low or dew risk. Expected outcome: By 02:30 MDT, reduce live VPD from 0.89 kPa to within 0.05 kPa of the active high edge and clear alert 10299 without creating VPD-low stress. Through 09:30, hold 0 dew-risk hours, no cold stress, no more than 0.25 h VPD-low stress, and avoid unnecessary wet/fog runtime once the corridor is recovered. Through 14:00, keep any renewed VPD-high error within 0.20 kPa of the crop-owned high edge and avoid another VENTILATE moisture-gap alert; accept physics- or storm-driven temperature deviations rather than chasing the target line inside the corridor.

Setpoints

Thursday August 13

Primary crop-band changes:

Tactical tunable changes:

Changed secondary parameters:

🌅 Morning Cycle (6:14 AM) — iris-20260813-0611

Reflection

Validating previous cycle: iris-20260812-2002

Previous hypothesis: SUNSET plan for Wednesday 2026-08-12. Today underperformed the seven-day baseline: score/compliance 62.4 versus 69.2, with VPD the slight legacy-axis bottleneck at 51.2% versus 51.8% temperature compliance. Heat stress slightly dominated at 7.69 h, followed by 7.22 h VPD-high, 0.31 h VPD-low, 0 cold stress, 8.3F minimum dew margin, and 0 dew-risk hours. Reliable electric/gas/water cost fields are unavailable, so exact cost attribution cannot be made; the context shows total water 168 gal versus a 276 gal seven-day average and runtime was cooling/fog/lighting-heavy rather than heating-heavy, but raw kWh is not used. I evaluated iris-20260812-0610 at 5/10, exactly matching the deterministic anchor of 5… Result: Plan governed about 06:14-20:06 MDT. Window score/compliance was 48.4, with legacy temperature compliance 42.1% and VPD compliance 45.2%. Stress was materially worse than the hypothesis implied: heat 4.85 h, VPD-high 3.53 h, VPD-low 0.64 h, cold 0 h. Conditions forecast grading: outdoor peak temperature matched (82.0F predicted, 82.2F actual); the dry-air forecast overcalled severity (39% predicted minimum RH versus 57% indoor proxy), while solar was severely undercalled (413 versus 1285 W/m2). Stress-window grading: the predicted medium VPD-high window occurred and was at least medium-to-high by duration (3.53 h); the predicted low heat window was a clear miss because heat dominated at 4.85 h; the high VPD-low window began after this plan was replaced and is unproven rather than failed. Rationale grading: band-relative mister engagement/all-zone thresholds and the shorter 30 s pulse gap preserved conditional authority but did not achieve good corridor performance under the unforecast solar spike; fog escalation to 0.20 was directionally appropriate for the 07:00-17:00 window but did not prevent 3.53 h VPD-high or 4.85 h heat; the 200 gal water cap was nonbinding because the day used 140 gal total/28 gal misting, so its expected anti-budget-gap effect was not tested; the 19:00 wet cutoff’s humid-night effect was outside this plan’s elapsed window. Dew safety did hold for the completed day (7.4F minimum margin, 0 dew-risk hours), and no cold stress occurred. Overall: accurate temperature peak and safe dew outcome, but poor 48.4 corridor score, major solar forecast miss, and underestimated heat severity. Score: 4/10

New finding: On convective/variable-cloud days, an accurate outdoor-temperature forecast does not validate the solar forecast: solar can exceed the predicted peak by more than 3x and create multi-hour heat/VPD-high stress. Keep morning thermal lead and moisture authority live-edge-triggered and band-coupled rather than suppressing them from a low solar forecast alone. → Added to Lessons Learned

Hypothesis

Testing: Test whether a forecast-bias-discounted, band-relative midday moisture surface activates only on an observed high-edge breach, limiting VPD-high stress and water use without carrying wet assist into the forecast humid night. Expected outcome: Through the next SUNRISE replacement, keep dew-risk hours at 0, VPD-low stress below 0.5 h, combined heat plus VPD-high stress below yesterday’s 19.35 h, and total water below the 276 gal seven-day average; no recurrence of a VENTILATE moisture-gap alert after any live midday VPD-high breach.

Conditions

Mostly overcast humid dawn, brief warm/dry midday pulse with historically overpredicted raw VPD, then a sharp 91-94% RH humid-night reversal. Fresh forecast solar conflicts with the earlier assembled calibration table, so live edge crossings govern activation.

Expected stress windows

Parameter rationale

Setpoints

Thursday August 13

Primary crop-band changes:

Tactical tunable changes:

Friday August 14

Primary crop-band changes:

Tactical tunable changes:

Changed secondary parameters:

🌆 Evening Cycle (8:06 PM) — iris-20260813-2001

Reflection

Validating previous cycle: iris-20260813-0611

Previous hypothesis: SUNRISE plan for Thursday 2026-08-13. Wednesday closed at score 58.8 versus the 69.2 seven-day average. Legacy temperature and VPD compliance were tied at 49.9%; heat stress slightly dominated at 10.00 h, followed by 9.35 h VPD-high, with 0 cold, 0.31 h VPD-low, minimum dew margin 8.3F, and 0 dew-risk hours. Reliable electric/gas/total-cost fields are unavailable. Water was 168 gal, including 43 gal misting, below the 276 gal seven-day average, so resource use was not the constraint. I evaluated iris-20260812-0610 at 5/10, exactly matching deterministic anchor 5. The formal evaluator reported zero guardrail events, while the assembled broader transition audit reports many holds/misses; this scope discrepancy is treated as safe dew outcome but uncertain materialization, not silently folded into the anchor. At 06:11 MDT the house is safe and inside its dispatcher-owned corridor at 72… Result: Plan governed approximately 20:06-00:19 MDT. Window score/compliance was 63.5; legacy temperature compliance was 45.4% and VPD compliance 34.3%. Stress was limited in absolute duration but mixed: 0.54 h heat, 0.52 h VPD-low, 0.06 h VPD-high, and 0 h cold. The live sunset VPD-high moisture gap largely recovered, so retaining band-coupled moisture authority until measured recovery was directionally successful. The subsequent 90-96% RH humid reversal also arrived, and tapering avoided dew danger: the completed day held a 7.4F minimum dew margin and 0 dew-risk hours. However, VPD-low reached 0.52 h during the short governed window and VPD compliance remained only 34.3%, so the humid-night handoff was not well centered inside the corridor. Reliable electric/gas/total-cost fields were unavailable; daily water was 140 gal, including 28 gal misting, versus a 274 gal seven-day average, so water was not binding. Overall the plan achieved recovery and safety but only moderate corridor performance. Score: 6/10

New finding: After an evening VPD-high recovery immediately before a very humid night, retain band-coupled wet authority until measured recovery, then taper promptly with a high dew-margin floor; judge success by both avoiding renewed VPD-high and limiting the resulting VPD-low window, not by moisture-gap closure alone. → Added to Lessons Learned

Hypothesis

Testing: Retain band-coupled moisture through observed evening recovery, then test whether a 12F dew-floor/high-churn-penalty humid-night posture prevents VPD-low and condensation while a forecast-discounted 08:00 handoff preserves authority for Friday’s ramp. Expected outcome: By 22:30 MDT, either measured VPD recovers inside the served high edge and alert 10302 clears, or the sticky VPD-high guardrail retains the evening assist surface. From midnight through 07:00, hold dew-risk hours at 0, minimum dew margin at or above 7F, VPD-low stress below 0.5 h, and cold stress at 0. The 08:00 fallback opens band-relative, forecast-discounted authority before Friday’s hot-dry ramp.

Conditions

Live VPD-high moisture gap at sunset, then 90-96% RH overnight and a forecast-discounted hot-dry Friday ramp.

Expected stress windows

Parameter rationale

Setpoints

Thursday August 13

Primary crop-band changes:

Tactical tunable changes:

Friday August 14

Primary crop-band changes:

Tactical tunable changes:

Changed secondary parameters:

End-of-Day Summary

Climate

Interior Light Evidence

• Crop DLI: unavailable
• Reason: source_reading_missing
• Provenance: legacy_invalid_exterior_proxy_plus_fixture_estimate
• Validity: dli-validity-v1; 2024-01-01T00:00:00+00:00 → open
• Qualified-light minutes and photoperiod remain independent of unavailable DLI.

Stress Hours

• Heat stress (>85°F): 5.5h
• VPD stress (>2.0 kPa): 10.7h
• Cold stress (<55°F): 6.3h

Economics

Equipment Runtimes

Water

• Total: 140 gal
• Mister: 28 gal

Crop Health (Gemini Vision)

Hourly Pattern

7-Day Stress Context