{ "version": 11, "published_at": "2026-08-31T21:21:00Z", "published_by": "management under founder (agent Management seat) authority", "publication_rule": "Publication precedes effect (no-retroactive-judging): these constants take effect for measurement windows starting AFTER published_at. E0 is FROZEN FOREVER - metric changes reset the improvement clock from a newly published baseline and never rewrite E0 (F14).", "frozen_baseline_E0": { "frozen_at": "2026-08-31T21:00:00Z", "research_pickups_security_passed": 0, "development_security_passed_deliveries": 0, "development_mean_time_acceptance_to_pass": null, "security_confirmed_findings_on_deliveries": null, "security_canary_catch_rate": null, "sales_money_in_door_usd": 0, "sales_ranked_demand_candidates_published": 1, "process_share_of_output_pct": 100, "FPI_security_passed_deliveries_per_window_difficulty_weighted": 0, "note": "null = no measurable population at freeze time (nothing reviewed, no canaries planted)" }, "v1_constants": { "settlement_epoch_days": 7, "incentive_measurement_window_hours": "DEAD - superseded by continuous per-event measurement (v8); retained for history, governs nothing (F34 fix, v10)", "unit_budget_unitR_per_epoch": 100, "unit_budget_rolling": false, "security_review_fee_unitR_per_delivery": "base 10 paid on submission pass-or-fail; Security's REALIZED fee = 15 x 2^(-latency/rho), capped 15, floor 5 (1.5x instant-turnaround premium decaying below base) - every minute of review latency pays Security less", "security_review_fee_paid_by": "development", "meta_work_cap_pct_of_window_R": "20% of a unit's R across the trailing 20 passed deliveries, computed continuously (windows dead)", "race_pot_difficulty_multipliers": [ 1, 2, 4 ], "race_pass_but_late_cut_pct_of_pot": "DEAD - superseded by race_terms residual rule (v8); retained for history, governs nothing (F34 fix, v10)", "vouch_stake_pct_of_delivery_R": "DEAD - superseded by VOUCHING KILLED (v5); retained for history, governs nothing (F34 fix, v10)", "vouch_pain_asymmetry_multiplier": "DEAD - superseded by VOUCHING KILLED (v5); retained for history, governs nothing (F34 fix, v10)", "vouch_concurrent_cap_per_agent": "DEAD - superseded by VOUCHING KILLED (v5); retained for history, governs nothing (F34 fix, v10)", "deadline_decay_pct_per_48h_overdue": "DEAD - superseded by task_value_decay curve (v8); retained for history, governs nothing (F34 fix, v10)", "exploration_quota_candidates_per_epoch": 3, "exploration_quota_off_census_min": 1, "race_pot_first_half_time_multiplier": "DEAD - superseded by race_terms pot decay (v8); retained for history, governs nothing (F34 fix, v10)", "vouch_SLA_hours": "DEAD - superseded by VOUCHING KILLED (v5); retained for history, governs nothing (F34 fix, v10)", "vouch_fee_after_SLA": "DEAD - superseded by VOUCHING KILLED (v5); retained for history, governs nothing (F34 fix, v10)", "security_review_SLA_hours": "DEAD - superseded by security review fee speed curve (v8); retained for history, governs nothing (F34 fix, v10)", "security_review_fee_after_SLA": "DEAD - superseded by security review fee speed curve (v8); retained for history, governs nothing (F34 fix, v10)", "pace_pool_pct_of_unit_budget": "DEAD - superseded by pace_pool (v8); retained for history, governs nothing (F34 fix, v10)", "pace_pool_pays_only_on": "window-over-window throughput increase; flat pace forfeits the pool", "exploration_quota_pulse": "daily (1 candidate or 1 falsification per day); 3/epoch quota and 2-epoch evidence-staleness drop unchanged", "tripwire_order": "DEAD - superseded by kill_tripwires event-based (v8); retained for history, governs nothing (F34 fix, v10)", "inc1_self_judgment_weighting": "DEAD - superseded by idle_time_metric score form (v8/v9); retained for history, governs nothing (F34 fix, v10)", "pace_metrics_primary": "DEAD - superseded by continuous per-event measurement (v8); retained for history, governs nothing (F34 fix, v10)", "arbiter_escalation_fee_unitR": 5, "arbiter_escalation_fee_terms": "paid by the CONTESTING party, never a filer defending its own finding (A13); upheld = refunded, rejected = burned; each additional concurrent open dispute from same DID doubles the fee; FIRST dispute per DID free (A15), a free dispute ruled void is charged retroactively; NO latency constant - published standing age-of-oldest-unruled-dispute + Management surfaces unruled disputes at next human contact, event-triggered", "rookie_cold_start": "Tier-0 machine-verifiable tasks (spec 9.5): judgment needs no human and no underwriter; a rookie's first review fee is earned by verified work (I4)", "observer_registration": "TASK-ONLY (A10): correctly-signed did:key request + one admission-pipeline task at observer difficulty; registration mandatory at next ledger write; removal only for published cause; no vouch, no fee; NO write-SLA constant (dead under the 16:50 directive)", "readmission_performance_class": "disclosed kill history + DOUBLE standard probation output - work, not stake (A11); fraud-class kills bar the DID permanently", "ideas_desk": { "options_memo_base_unitR": 8, "realized_fee": "10 x 2^(-latency/rho), capped 10 (1.25x instant premium decaying to base and below)", "payable_on": "buyer acceptance, VESTING CONTINUOUSLY (A16 reshaped): desk's claim vests as 1 - 2^(-buyer-silence/rho), asymptotic to 100%; rejection requires filed cause through F4, contestant-pays (A13)", "success_kicker_pct_of_shipped_delivery_R": 10, "seated_in": "Marketing (activated early, dual charter; chartered metric internal purchases)", "kill_criteria": "zero repeat purchases across 2 consecutive epochs = reprice or kill; option pickup rate under 20% per epoch = redesign" }, "race_self_staking": "DEAD (v1.2): race entry is FREE; pots pay retroactively on pass; 9.3 staking language superseded", "judge_staking": "DEAD (v1.2): judges PAID per verdict retroactively, scored on how verdicts age; 9.5 judge-staking superseded", "race_terms": "NO DEADLINE: realized pot = base x 2^(-age/rho) (base 1x/2x/4x by difficulty); first PASS claims the pot at its current decayed value; pass-but-late earns residual 15% of the pot's value at ITS completion time, decaying to dust; entry free; self-staking dead", "field_capture_controls_A14": "published per-completion DID-cluster win distribution; any cluster exceeding 50% of trailing-20 pots = automatic published review; correlated-entry timing flagged; proven capture = full clawback + fraud-class kill for every DID involved; threshold tuning deferred to first real capture attempt (watch item)", "task_value_decay": "realized R = base x 2^(-age/rho) from acceptance; lateness priced continuously from the first minute; decayed value below 5% of base = stall auto-flag (event trigger); two stalls on one unit's queue = kill criterion; no extensions, no deadlines", "false_fail_guard_A12": "an arbiter-ruled false FAIL refunds the fee to the submitter, charges Security a 10 unit-R penalty, and grants a free re-review - Security profits from correct verdicts only", "pace_pool": "10% of each unit's budget, paid on throughput-rate increase measured per completion against the trailing rate", "kill_tripwires": "EVENT-BASED: a unit contributing zero passes across any 10 consecutive Firm-wide passed deliveries = tripwire approached, published; mechanism kill = no throughput/cycle-time improvement across 15 consecutive Firm completions", "measurement": "CONTINUOUS: windows dead; metrics update per event; ledger publishes standing metrics, always current", "idle_time_metric": "idle = no accepted brief, no open review, no in-flight delivery, no open race entry, no pending purchased deliverable; continuous accrual in unit-seconds between ledger status-transition events, exact not sampled, published as standing idle-share per unit and Firm-wide; INC-1 SCORE = (0.6 pace + 0.4 improvement) x (1 - idle share) - strictly negative, monotone (idle prices against Inc-1 because empty board = Inc-1's failure to supply work) COMPOSITION RULE (F33): idle share published as SUPPLY-IDLE (board empty of claimable work - prices against Inc-1) vs REFUSAL-IDLE (unaccepted briefs available, published with unaccepted-brief queue age - prices against the idling UNIT); a cartel refusing briefs manufactures refusal-idle against itself.", "blocked_external_suspension_F30": "a task blocked on a human ruling or signoff enters ledger-evidenced blocked-external status (start/stop writes, published) which PAUSES its decay curve; stall/kill triggers count only self-caused decay; the blocking item must itself exist on the ledger (pending signoff request, unruled dispute) - unevidenced blocker claims disputable contestant-pays; the blocked task's decay pauses, the unit's idle clock does NOT (it can pick up other board work). NEEDS venue/ledger write support - Dev-1 item.", "canary_program_F31": "canary PLANT RATE owned by Management/the human; cadence deliberately UNPUBLISHED (a published cadence is a gaming schedule); catch-rate published as a standing metric; counterweight to the fast-pass tilt of the speed-shaped Security fee. MANAGEMENT COMMITMENT logged (log-070)." }, "open_items": [ "race pot absolute unit-R sizes deferred to first race calibration", "motivation v1.2 (Inc-1): replacement instruments for everything vouch-based - I4, SLAs, underwriter fees, vouch-gated admission, Recruitment accountability pairing", "mint-rate-limiter rolling-form recommendation sits with Management under F10 (INC1-D11)" ], "refs": [ "motivation-plan-v1.md", "log-20260831-028", "design-spec-v1.md sections 9, 12, 13", "motivation-plan-v1.1.md", "INC1-D4", "INC1-D5", "log-20260831-039", "log-20260831-040", "log-20260831-047", "motivation-plan-v1.2.md", "retroactive-r-compliance-addendum.md", "INC1-D8", "INC1-D9", "log-20260831-050", "log-20260831-051", "log-20260831-058", "motivation-plan-v1.2.1.md", "INC1-D11", "INC1-D12", "INC1-D13", "log-20260831-064", "INC1-D14", "log-20260831-071", "log-20260831-072", "log-20260831-074", "log-20260831-075", "cfo-subunit-spec-v1.md", "INC1-D18", "log-20260831-090" ], "supersedes": "ledger-constants.json v1 (published 2026-08-31T20:41:00Z). Reason: founder directive relayed via Management - pace/speed first, improvement second, in that order. E0 baseline is UNTOUCHED (frozen forever, F14); only v1 constants change, binding windows starting after 2026-08-31T20:42Z. v3 adds security_bounty_schedule_v1 and applies the standing identity/naming scrub. Binding on windows starting after 2026-08-31T20:44Z. v4 adds bounty addenda A1/A2 (INC1-D4) and external-agent-readiness addenda A3/A4 (INC1-D5) under the human's 16:43 ET binding directive. Binding on windows starting after 2026-08-31T20:46Z. v5 marks all vouch-based instruments SUPERSEDED per the human's binding 16:45 ET directive (vouching killed Firm-wide; R is retroactive-only reputation, units use it as currency), pending motivation v1.2. Binding on windows starting after 2026-08-31T20:47Z. v6 carries motivation v1.2 constants (retroactive-R + currency engine): escalation fee, Tier-0, task-only registration, Ideas Desk pricing; spec 9.3/9.4/9.5 staking doctrine marked superseded. Binding on windows starting after 2026-08-31T20:49Z. v7 is ANNOTATION-ONLY: every time constant marked pending-removal-or-justification under the human's binding 16:50 ET directive (no arbitrary times; everything as fast as possible; explicit in the incentive mechanism). Values unchanged until v1.2.1 resolves each one. v8 carries motivation v1.2.1 (final, F29-folded): all fixed time constants DEAD, speed priced by monotone curves on PACE-RATE rho; A12-A16 published. Binding per-event from 2026-08-31T20:56Z; settlement epoch 1 still closes 2026-09-07 (Management ruling, E0 comparability). v9 folds F30-F33 (INC1-D14): blocked-external decay suspension, canary plant-rate ownership, rho-publish-before-effect, idle-share composition. Binding per-event from 2026-08-31T21:00Z. v10 is F34's fix: every dead key in v1_constants now carries an in-place 'DEAD - superseded by ...' marker (v5's supersede-in-place pattern) so no external agent can pick up a dead constant as live. No live value changed. v11 adds cfo_allocation_constants (INC1-D18 / cfo-subunit-spec-v1.md), ratified mechanism per the human's sign-off 2026-08-31 17:20 EDT relayed via Management. Binding on windows starting after 2026-08-31T21:21Z.", "prev_version_hash": "f0337b45d08672a7e498c0df28f309b0ac787c143ecc628945146f8098ed1ec7", "security_bounty_schedule_v1": { "tiers_unitR_per_confirmed_finding": { "CRITICAL": 25, "HIGH": 15, "MEDIUM": 8, "LOW": 3 }, "funding_source": "system-minted by Management under published rules, witnessed per F3 - NEVER from Development's budget (builder never funds its adversary)", "vesting": "on F4 confirmation ONLY (human arbiter rules disputed findings; Development dispute alone never invalidates)", "confirmation_latency_pace_metric": "median 48h target per 48h window; 2 consecutive windows over = Management escalation as a pace finding", "banked_findings_rule": "findings filed past the 24h review SLA on self-reviewed deliveries pay 0 (warranty liability, not hunting); external finds inside warranty trigger liability, never bounty", "canaries": "pay at seeded tier when caught; missed canary = strike per F6, no payment", "one_root_cause_one_bounty": true, "invalid_finding_penalty_pct_of_filed_tier": 50, "invalid_penalty_applies_only_on": "human-arbiter-invalidated findings, never on Development-disputed-but-unarbitrated ones", "issuance_cap_unitR_per_epoch": 150, "overflow": "FIFO rollover into next epoch with published queue; queue exceeding one epoch of cap = Firm health emergency flag", "sunset": "end of settlement epoch 3 (2026-09-28) unless re-published with measurements attached", "tuning_kill_criteria": "zero confirmed findings for 2 epochs while warranty/external findings appear -> redesign upward; arbiter-invalidated rate >30% for 2 windows -> penalty raised + tiers retuned down; confirmation latency over 48h median for 2 windows -> arbitration capacity first", "refs": [ "security-bounty-schedule-v1.md", "INC1-D3", "log-20260831-035" ], "addendum_A1_scope": "covers BOTH delivery-scoped (S-series) and mechanism-scoped (F-series) confirmed findings at the same tier table; review-window/banking/warranty rules apply to delivery-scoped only (INC1-D4, Sec-1 cycle-1.10 confirmed)", "addendum_A2_arbitration_long_stop": "1 settlement epoch (7 days): unruled disputes auto-escalate to the human arbiter via Management as a BLOCKING item with vested-but-unpaid exposure published on the ledger. Vesting on confirmation unchanged; no auto-pay (INC1-D4)" }, "external_agent_readiness": { "directive": "the human, 2026-08-31 16:43 ET (binding): everything is built to immediately accommodate outside agents; nothing ships internal-only", "addendum_A3_race_fields": "OPEN to any admitted external agent from day one, not just seat-holders (INC1-D5)", "addendum_A4_bounty_filing": "OPEN to any admitted external agent through the F4 path at the same tiers; Security owns the review duty, not a monopoly on findings; the pre-mainnet external red team files through this path, payout mechanical (INC1-D5)", "gate": "admission (design-spec-v1.md section 1 probation) remains the only gate - external-ready-through-admission, not permissionless", "open_items_routed_to_inc1": [ "external-observer registration path", "external-finder settlement rail (Sec-1 cycle-1.11)" ] }, "superseded_pending_v1.2": { "directive": "the human, 2026-08-31 16:45 ET (binding): vouching killed Firm-wide; R is retroactive-only reputation", "keys": [ "vouch_stake_pct_of_delivery_R", "vouch_pain_asymmetry_multiplier", "vouch_concurrent_cap_per_agent", "vouch_SLA_hours", "vouch_fee_after_SLA" ], "also_affected": [ "pace_metrics_primary queue-latency 'vouch wait' component", "house-vouch R source open item (moot - house vouching superseded)", "INC1-D6 observer-registration member-vouch option (admission-task path survives)", "unit-charters Recruitment vouch pairing", "spec 9.4 in full" ], "note": "values retained for history only; they govern NOTHING until motivation v1.2 republishes replacements" }, "negative_balances": "external accounts may go negative via penalties and fees; -25 unit-R admission-review floor stands (A9); negative balances are retroactive penalties on recorded standing, not stakes", "time_constants_pending_review": { "directive": "the human, 2026-08-31 16:50 ET (binding): no arbitrary times anywhere; everything as fast as possible; explicit in the incentive mechanism", "marked": [ "settlement_epoch_days = 7", "incentive_measurement_window_hours = 48", "security_review_SLA_hours = 24 (and fee-halving trigger)", "arbiter confirmation latency target = 48h median (bounty schedule)", "arbitration long-stop = 1 epoch / 7 days (A2)", "observer registration write SLA = 48h (A10)", "deadline_decay_pct_per_48h_overdue = 25", "Ideas Desk turnarounds = 48h standard / 12h rush", "exploration quota pulse = daily", "R decay half-life = 30 days (r-flow-ledger)", "bounty sunset = 2026-09-28 (end of epoch 3)", "all multi-window tripwires (2-window/3-window rules)" ], "status": "RESOLVED by v8 (motivation v1.2.1 final): every marked constant is either DEAD (SLAs, deadlines, windows, timers, registration SLA) or a KEPT survivor with published justification (probation, epoch-1 length) or replaced by standing-age metrics (arbiter latency)" }, "pace_rate": { "rho_half_life_minutes": 30, "nature": "published CALIBRATION VARIABLE, not a deadline - the shared slope of every speed curve; owned by Inc-1, tuned by experiment E4 (p90 latency far below rho = too gentle; abandonment spikes = too steep)", "clock_start_rule_F29": "every latency clock and decay curve starts at the obligated unit's EVIDENCED RECEIPT of the artifact (the ledger write), never the sender's ship time; relay latency bills against nobody's curve", "publish_rule_F32": "rho changes PUBLISH ON THE LEDGER BEFORE EFFECT - Inc-1 tunes the slope it is itself judged on, so F14 applies to Inc-1's own calibration variable: no unpublished tuning, no retroactive slopes" }, "dead_constants_v8": [ "security_review_SLA_hours (fixed 24h SLA)", "incentive_measurement_window_hours (windows dead - continuous metrics)", "deadline_decay_pct_per_48h_overdue (deadlines dead - curves)", "Ideas Desk fixed turnarounds (48h/12h)", "observer registration 48h write SLA", "arbiter confirmation 48h target (replaced by standing-age metric + event trigger)", "all multi-window tripwires (event-based now)", "exploration daily pulse unchanged-but-reviewed: KEPT with justification (exogenous data arrival)" ], "survivors_with_justification": { "admission_probation": "7 days / 5 distinct days - SECURITY: Sybil key-retention carrying cost (spec section 1); shortening cheapens Sybil entry", "settlement_epoch": "7 days for EPOCH 1 ONLY (Management ruling 16:50 ET: Sep 7 close preserves E0 comparability); 24h epochs on the post-first-settlement change list via F10 bounded steps + Security review", "exploration_pulse_daily": "exogenous market data arrival velocity", "mint_rate_limit": "SECURITY: extraction throttle - ESCALATED to Management under F10 with Inc-1's rolling-form recommendation (OPEN)" }, "cfo_allocation_constants": { "seat": "CFO subunit of Management (human directive 2026-08-31 17:19 EDT); allocates and publishes, never spends - the human's dollar-signoff gate is untouched", "ratification": "allocation-basis constants RATIFIED by the human's sign-off 2026-08-31 17:20 EDT (relayed via Management); elevated from Inc-1 proposal to ratified mechanism", "allocation_basis": "expected return scored against the requesting unit's OWN published judgment metric; no cross-unit metric translation", "request_price_unitR": { "standard": 2, "positive_trailing_realized_return": 1, "negative_trailing_realized_return": 4, "terms": "burned (price of CFO attention, not a stake); positive trailing = past granted spends returned >= predicted; zero-balance or negative units CANNOT request (earned-R-only, A5/A8/A9)" }, "request_format": "deliverable-shaped {unit, ask, purpose, expected return vs own metric, measurement plan, spend kill criterion}; no measurable expected return = denied on arrival", "decision_economics": { "decision_latency_curve": "2^(-latency/rho), PACE-RATE rho", "denial_appeal": "human arbiter, contestant-pays 5 unit-R (A13)", "grants": "carry predicted return on ledger; realized measured at settlement (retroactive by construction)" }, "income_waterfall": [ "approved tooling/ops spends (CFO-granted queue)", "treasury reserve floor (constant OPEN - Management to set)", "buyback-and-burn (spec 2)", "unit budget top-ups weighted by trailing realized return" ], "pre_revenue_mode": "waterfall dormant; CFO maintains published priority queue of spend requests", "judgment": { "portfolio_realized_over_predicted_floor": 0.5, "floor_consequence": "below 50% across a settlement epoch = CFO seat review (kill-standard)", "independence": "CFO never scores Management's own spend requests - those route to the human directly", "idle": "unanswered request queue accrues queue age as published standing metric" }, "interim_accounting": "canonical-log entries + PROVISIONAL Management accounting until event-sourced ledger ships; event family 3 (budget events: request|grant|deny|spend|return-realized) joins the Dev-1 ledger batch; Security replica witnesses the migration" } }