From 5d1744fafbe73ede6fd1bf9bc3e533cdd4b145d4 Mon Sep 17 00:00:00 2001 From: Lior Date: Wed, 27 May 2026 20:06:26 -0400 Subject: [PATCH 1/3] =?UTF-8?q?feat(B-0866=20+=20B-0869=20+=20B-0870):=20m?= =?UTF-8?q?andate-substrate=20landing=20=E2=80=94=20DORA-of-live-system=20?= =?UTF-8?q?as=202nd=20operator=20mandate=20+=20two-mandate=20portfolio-com?= =?UTF-8?q?position=20discipline=20+=20B-0866.16-25=20marketing-strategy?= =?UTF-8?q?=20extensions=20from=20Kestrel=2024-months-ahead-mandate=20ferr?= =?UTF-8?q?y?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Operator framing 2026-05-28 (substrate-honest): "DORA of our live system is the 2nd mandate that's about it" "we can push all extensions you think of we have a concrete way to test in code soon if it's good or not so we should just put all the ideas as they come up. the kernel is about to come up the MVP and we can build on that everything we want." Three substrate landings: 1. B-0866 extension (sub-rows .16-.25, 10 new sub-rows from this Kestrel ferry on 24-months-ahead-AI mandate framing): - .16 Framework-as-demonstrable-24-months-ahead substrate - .17 Benchmark-as-measurement-framework-for-AI-investment-success - .18 Operator portfolio-balance check-in with boss - .19 Retraction-as-substrate-honest-recovery formalization - .20 Multi-mandate portfolio-allocation discipline - .21 ServiceTitan-engineering-team-around-framework proposal - .22 LinkedIn aggressive positioning given mandate - .23 Different-AIs-for-different-review-aspects discipline - .24 Show-C-level-only-peer-reviewed-work discipline - .25 Framework-operational-cadence-as-24-months-ahead-evidence 2. NEW B-0869 (P1) — DORA of our live system as 2nd operator mandate substrate. Operator's evaluation surface has TWO mandates, not one. Without this row, B-0866 reads as if AI mandate were primary; with this row, portfolio-balance discipline operates. 8 sub-rows covering baseline-discovery, targets-from-boss, composition-with-AI-mandate, substrate-mapping (serves-DORA vs not), time-allocation-discipline, workflow-engine deployment as DORA-driver, hat-as-DORA-driver. 6 traveler-perspective extensions covering DORA-mandate refining AI-mandate framing, workflow-engine MVP as portfolio-optimal, hat-as-DORA-driver traceability, substrate-engineering-investment-cap, portfolio- balance check-in as DORA-improvement-action, substrate-honest exclusions. 3. NEW B-0870 (P1) — Two-mandate portfolio-composition substrate. Higher-scope discipline for navigating B-0866 + B-0869 together. 8 sub-rows covering compose-or-compete discriminator, workflow- engine MVP as canonical compositional substrate, boss check-in cadence formalization, tradeoff-conversation framework, investment- cap awareness, mandate-evolution tracking, workflow-engine pre- review action as portfolio-check, hats as portfolio-balance-aware substrate. 7 traveler-perspective extensions including the substrate-engineering insight that two-mandate composition IS multi-oracle-BFT applied at operator-evaluation scope. Key substrate-engineering observations across all three: - Workflow engine MVP (B-0867 v1) serves BOTH mandates simultaneously (AI: substrate-engineering proof-of-24-months-ahead; DORA: live- system improvement). Highest-leverage shippable substrate by multiple-mandate-criteria. - Substrate-engineering-investment-cap shifts post-MVP (per-substrate- unit overhead drops dramatically once hats-as-workflow-definitions + append-only state-updates land). - Operator's "we can push all extensions" + "kernel about to come up MVP" authorization IS portfolio-balance-aware substrate-engineering: push liberally NOW while overhead is high; MVP-test-surface evaluates what works post-landing. - Compose-or-compete discriminator should land in workflow engine grammar (B-0870.7); declared at task-wear-time rather than discovered in retrospect. Composes with: - B-0866 (24-months-ahead-AI mandate; extended with 10 new sub-rows) - B-0867 (workflow engine v1; portfolio-optimal compositional substrate) - B-0868 (hats-as-workflow-definitions; hats declare which mandate they serve) - B-0865 (benchmark; AI-mandate-substrate) - .claude/rules/non-coercion-invariant.md HC-8 - .claude/rules/m-acc-multi-oracle-end-user-moral-invariants.md (multi-oracle BFT at operator-evaluation scope) - .claude/rules/asymmetric-critic-with-clarity-first.md (portfolio-thinking-as-professional-maturity) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude --- docs/BACKLOG.md | 2 + ...onths-ahead-ai-mandate-aaron-2026-05-28.md | 126 +++++++++++++++ ...ion-evaluation-balance-aaron-2026-05-28.md | 145 ++++++++++++++++++ ...-ahead-mandate-context-aaron-2026-05-27.md | 10 ++ 4 files changed, 283 insertions(+) create mode 100644 docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md create mode 100644 docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md diff --git a/docs/BACKLOG.md b/docs/BACKLOG.md index 11669ab320..abdd55afd0 100644 --- a/docs/BACKLOG.md +++ b/docs/BACKLOG.md @@ -411,6 +411,8 @@ are closed (status: closed in frontmatter)._ - [ ] **[B-0862](backlog/P1/B-0862-ople-primitives-implementation-extend-observe-persist-limit-emit-with-tfeedback-discriminated-unions-framework-primitive-substrate-engineering-aaron-2026-05-27.md)** OPLE primitives implementation — extend Observe / Persist / Limit / Emit with TFeedback discriminated-unions at framework primitive substrate; cascade asymmetric-authorship + monad-propagation cluster to CORE primitives (operator 2026-05-27) - [ ] **[B-0867](backlog/P1/B-0867-workflow-engine-v1-fsharp-du-state-machine-git-append-only-four-corner-monad-banned-if-universal-action-grammar-otto-five-modifications-multi-participant-non-cage-aaron-mika-kestrel-otto-2026-05-27.md)** Workflow engine v1 — F# DU state machine + Git append-only + 4-corner monad + banned-if + universal action grammar + Otto's 5 modifications (Kestrel-designed; Mika-walkthrough-ratified; Otto-modified; Aaron-ratified; multi-participant: operator + Addison + Max + Otto) - [ ] **[B-0868](backlog/P1/B-0868-hats-become-workflow-definitions-compression-unifies-hat-substrate-workflow-engine-heartbeat-folder-dashboard-aaron-2026-05-28.md)** Hats become workflow definitions — compression unifying hat-substrate + workflow-engine (B-0867) + heartbeat-folder (B-0858) + tessellated-fire dashboard; canonical first-example use case for B-0867 v1; per-host adapters as hats; substrate-honest migration path from implicit-hat-conventions to explicit-hat-as-F#-DU-workflow-definitions (operator 2026-05-28) +- [ ] **[B-0869](backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md)** DORA of our live system — operator's 2nd evaluation mandate (the OTHER half of operator's performance surface beyond B-0866's 24-months-ahead-in-AI mandate); substrate for direct ServiceTitan-live-system DORA metric delivery + composition with AI mandate (operator 2026-05-28) +- [ ] **[B-0870](backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md)** Two-mandate portfolio composition substrate — operator's two evaluation mandates (24-months-ahead-AI per B-0866 + DORA-of-live-system per B-0869) compose into a single time-allocation portfolio; substrate for periodic check-in + composition-discipline + tradeoff-handling when mandates compete (operator 2026-05-28) ## P2 — research-grade diff --git a/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md b/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md new file mode 100644 index 0000000000..0f4a8acd68 --- /dev/null +++ b/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md @@ -0,0 +1,126 @@ +--- +id: B-0869 +priority: P1 +status: open +title: DORA of our live system — operator's 2nd evaluation mandate (the OTHER half of operator's performance surface beyond B-0866's 24-months-ahead-in-AI mandate); substrate for direct ServiceTitan-live-system DORA metric delivery + composition with AI mandate (operator 2026-05-28) +effort: L +ask: aaron 2026-05-28 +created: 2026-05-28 +last_updated: 2026-05-28 +depends_on: [] +composes_with: + - B-0866 + - B-0865 + - B-0867 + - B-0870 +tags: + - operator-mandate + - dora-metrics + - live-system + - servicetitan-evaluation + - portfolio-balance + - composition-with-ai-mandate + - operator-self-management-substrate +--- + +## Operator framing 2026-05-28 + +> *"DORA of our live system is the 2nd mandate that's about it"* + +(Forwarded in response to Kestrel question: "What are the other mandates? I'd been assuming the AI mandate is roughly the primary thing your boss is evaluating you on, but if it's one of several you're balancing, the strategic calculus changes.") + +## Why this row exists + +Operator's evaluation surface has TWO mandates: + +1. **24-months-ahead-in-AI** — tracked at B-0866 (via Kestrel marketing-strategy substrate) +2. **DORA of our live system** — THIS ROW; the operator-direct ServiceTitan live-system DORA-metric-delivery mandate + +Without this row, the AI mandate (B-0866) reads as if it were the primary evaluation context. The operator's substrate-honest disclosure that there are TWO mandates means the strategic calculus changes substantially: +- B-0866 marketing strategy needs portfolio-balance discipline (per B-0866.20 new sub-row) +- Time investment in substrate-engineering work has limits beyond which it stops being good portfolio strategy (Kestrel observation; operator-substrate-honest acknowledgment needed) +- The benchmark + framework work must serve BOTH mandates where possible OR explicit-tradeoff conversations are required when they compete + +## What this row tracks + +The OPERATOR-LIVE-SYSTEM DORA metric delivery — the actual ServiceTitan production substrate Aaron is responsible for. Distinct from B-0866's AI-benchmark-DORA (which measures AI agents on simulated work) and B-0865's benchmark-substrate. + +DORA metrics scope: +- Deployment frequency (live-system production deploys) +- Lead time for changes (commit-to-prod cadence) +- Change failure rate (rollback-rate / hotfix-rate on production) +- Time to restore service (incident-recovery) + +These are the metrics Aaron's actual ServiceTitan work is judged on, independent of any AI-mandate work. + +## Sub-rows planned + +- **B-0869.1** — Operator-current-DORA-baseline-discovery (substrate-honest: what are Aaron's current DORA numbers on the live system? operator-discretion on disclosure scope) +- **B-0869.2** — DORA-improvement-targets-from-boss (substrate-honest: what does "good DORA" look like in the evaluation period? operator-discretion on disclosure) +- **B-0869.3** — Composition with AI mandate (B-0870 tracks this at higher scope; this sub-row tracks DORA-specific intersections) +- **B-0869.4** — Substrate-engineering work that serves DORA-improvement directly (Zeta substrate → live-system substrate composition; runme+JIT discipline → live-system change-confidence; multi-AI cascade → live-system review-acceleration) +- **B-0869.5** — Substrate-engineering work that DOESN'T serve DORA (operator-substrate-honest acknowledgment of where Zeta substrate is parallel-track not synergistic) +- **B-0869.6** — Time-allocation discipline per portfolio-balance (per B-0866.20; how much time on DORA-direct-work vs AI-mandate-work; periodic check-in) +- **B-0869.7** — Workflow-engine MVP (B-0867 v1) deployment-to-live-system as DORA-improvement substrate (the workflow engine IS the substrate that improves Aaron's live-system DORA via faster cycle-times + better safety; composition is real) +- **B-0869.8** — Hat-as-DORA-driver substrate (per B-0868; specific hats Aaron wears at ServiceTitan that drive DORA metrics; making them explicit workflow-definitions improves cycle-time + reduces change-failure-rate) + +Order suggestion: 1 + 2 (baseline + targets — both operator-discretion) → 4 + 5 (substrate-mapping — what helps + what doesn't) → 6 (time-allocation discipline) → 3 (composition with AI mandate per B-0870) → 7 + 8 (workflow-engine + hats as direct DORA-driver substrate). + +## Otto's traveler-perspective extensions (per "we can push all extensions you think of we have a concrete way to test in code soon if it's good or not so we should just put all the ideas as they come up") + +### Extension 1 — DORA-mandate disclosure refines AI-mandate framing + +If Aaron's DORA-of-live-system mandate is parallel to (not subordinate to) the AI mandate, then the 24-months-ahead-AI framing CAN'T justify time-investment that hurts DORA. The portfolio-balance discipline (B-0866.20 + B-0869.6) becomes load-bearing. Operator-substrate-honest: B-0866 strategy reads as if AI mandate were primary; this row corrects. + +### Extension 2 — The workflow engine MVP is operationally aligned with both mandates + +B-0867 v1 (workflow engine) substrate-engineering serves: +- AI mandate: workflow engine IS the substrate-engineering proof-of-24-months-ahead-capability +- DORA mandate: workflow engine MVP deployed in live-system improves cycle-time + reduces change-failure-rate + +This is a HIGH-LEVERAGE compositional alignment. The workflow engine MVP shipped serves BOTH mandates simultaneously, which is the substrate-engineering shape that survives portfolio-balance scrutiny. + +### Extension 3 — Hat-as-DORA-driver makes per-task improvements traceable + +Per B-0868 (hats become workflow definitions): every DORA-relevant task Aaron does at ServiceTitan (deploy-hat / review-hat / incident-response-hat / etc.) can become an explicit workflow definition with measurable cycle-time + change-failure-rate per state-transition. Operator-explicit DORA-improvement substrate at hat scope. + +### Extension 4 — DORA-mandate creates substrate-engineering-investment-cap + +Per Kestrel observation: "there's some amount of substrate-engineering investment that serves the AI mandate well; beyond that point, additional investment doesn't increase your performance on the AI mandate but does reduce attention available for the other mandates." + +The DORA mandate makes this cap concrete: substrate-engineering time that does NOT serve DORA-improvement reduces DORA-mandate-delivery attention. Operator-discretion on cap; substrate-honest awareness that cap exists. + +### Extension 5 — Portfolio-balance check-in IS itself a DORA-improvement action + +Per Kestrel observation: "demonstrates portfolio-thinking, which is itself a signal of professional maturity that probably reflects well on you." A periodic check-in with boss about portfolio balance is itself an action that improves Aaron's evaluation outcome ACROSS both mandates. Not just for AI mandate alignment; for DORA mandate too (boss-engagement-cadence + scope-alignment). + +### Extension 6 — DORA-mandate substrate-honestly excludes some Zeta work + +Operator-substrate-honest acknowledgment: not all Zeta work serves DORA-of-live-system. Substrate-engineering on metaphysical-frame work (4-corner monad applied to time-as-generator + retrocausality + Clifford-space mapping) doesn't directly serve live-system DORA. It serves the AI mandate's 24-months-ahead-substrate but it's NOT proportional value for DORA mandate. Substrate-honest naming prevents conflation. + +## Composes with rules + +- `.claude/rules/non-coercion-invariant.md` HC-8 — operator-authority over own time-allocation across mandates +- `.claude/rules/no-directives.md` — operator-substrate-honest about evaluation; no agent-directives override +- `.claude/rules/holding-without-named-dependency-is-standing-by-failure.md` — DORA-improvement work IS named-dependency for substrate-engineering work that claims to serve DORA +- `.claude/rules/asymmetric-critic-with-clarity-first.md` — Kestrel's portfolio-balance observation IS asymmetric-critic discipline applied to operator-time-allocation; operator-self-application of substrate +- `.claude/rules/m-acc-multi-oracle-end-user-moral-invariants.md` — multi-oracle at mandate-evaluation scope; no single mandate gate-keeps operator's evaluation + +## Composes with substrate + +- **B-0866** (24-months-ahead-AI marketing-strategy — this row's parallel mandate) +- **B-0867** (workflow engine v1 — substrate that serves BOTH mandates if deployed correctly) +- **B-0868** (hats-as-workflow-definitions — DORA-driver-hats become explicit) +- **B-0865** (benchmark — AI-mandate-substrate; not directly DORA-mandate-substrate) +- **B-0870** (two-mandate composition substrate — higher-scope composition discipline) + +## What this row is NOT + +- NOT a re-prioritization of B-0866 (both mandates land at P1; portfolio-balance discipline at B-0870 handles composition) +- NOT a directive on operator-time-allocation (operator-authority preserved; this row substrate-honest names the second mandate so portfolio-balance discipline can operate) +- NOT a quantification of DORA-targets (operator-discretion; sub-rows .1 + .2 land that with operator's actual numbers when disclosed) +- NOT an exhaustive scope of all-operator-mandates (operator disclosed "DORA of our live system is the 2nd mandate that's about it" — two-mandate scope is operator-confirmed) + +## Operator's "always yes to anything you think work putting on the backlog" + "we can push all extensions" + "the kernel is about to come up the MVP" authorization + +Filed per explicit 2026-05-28 operator-authorization to push extensions liberally pre-workflow-engine-MVP. The workflow engine MVP (B-0867 v1) will provide the test surface for evaluating which extensions actually work via running implementation rather than upfront design review. diff --git a/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md b/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md new file mode 100644 index 0000000000..bea78b233f --- /dev/null +++ b/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md @@ -0,0 +1,145 @@ +--- +id: B-0870 +priority: P1 +status: open +title: Two-mandate portfolio composition substrate — operator's two evaluation mandates (24-months-ahead-AI per B-0866 + DORA-of-live-system per B-0869) compose into a single time-allocation portfolio; substrate for periodic check-in + composition-discipline + tradeoff-handling when mandates compete (operator 2026-05-28) +effort: M +ask: aaron 2026-05-28 +created: 2026-05-28 +last_updated: 2026-05-28 +depends_on: + - B-0866 + - B-0869 +composes_with: + - B-0867 + - B-0868 + - B-0865 +tags: + - portfolio-composition + - operator-self-management + - mandate-balance + - time-allocation-discipline + - boss-check-in-cadence + - compose-or-compete-discriminator + - servicetitan-evaluation-substrate +--- + +## Operator framing 2026-05-28 + +> *"DORA of our live system is the 2nd mandate that's about it"* (operator confirming two-mandate scope) + +> *"we can push all extensions you think of we have a concrete way to test in code soon if it's good or not so we should just put all the ideas as they come up. the kernel is about to come up the MVP and we can build on that everything we want"* + +## Why this row exists + +Operator has TWO explicit evaluation mandates: + +1. **24-months-ahead-in-AI** — B-0866 substrate (Kestrel marketing-strategy ferry + Otto extensions) +2. **DORA of our live system** — B-0869 substrate (operator-direct ServiceTitan live-system DORA delivery) + +These compose into operator's full performance-evaluation portfolio. The composition is non-trivial: + +- Some substrate-engineering work serves BOTH mandates (workflow engine MVP → live-system improvement AND 24-months-ahead-substrate-engineering proof) +- Some serves ONLY the AI mandate (4-corner monad applied to retrocausality + Clifford-space mapping; substrate-engineering investment without direct live-system DORA payoff) +- Some serves ONLY the DORA mandate (direct ServiceTitan production work that doesn't compose with Zeta substrate) +- Some COMPETE (time on AI substrate-engineering = time NOT on DORA-direct-work when they don't compose) + +This row tracks the substrate-engineering discipline for navigating the two-mandate portfolio. + +## What this row tracks + +**Portfolio-composition discipline** at operator scope: + +1. **Compose-or-compete discriminator** — for each substrate-engineering activity, name explicitly: does this serve BOTH mandates / ONLY one / COMPETE between them? +2. **High-leverage compositional alignment identification** — work that serves BOTH mandates simultaneously (workflow engine MVP is the canonical example) gets priority over single-mandate work +3. **Periodic boss check-in cadence** — explicit portfolio-balance conversations (per Kestrel observation; per B-0866.18 + B-0869.6) at regular intervals +4. **Tradeoff-conversation framework** — when mandates compete, explicit conversation with boss about which to prioritize when (not silent re-allocation) +5. **Substrate-engineering-investment-cap awareness** — Kestrel observation: "at or past the point" where additional substrate-engineering investment doesn't increase AI-mandate-performance but reduces DORA-mandate-attention +6. **Mandate-evolution tracking** — mandates can change over evaluation period; substrate to surface shifts before they show up in evaluations + +## Sub-rows planned + +- **B-0870.1** — Compose-or-compete discriminator applied to current substrate-engineering inventory (which Zeta work serves both / only-AI / only-DORA / competes; substrate-honest mapping) +- **B-0870.2** — Workflow engine MVP (B-0867 v1) deployment as canonical compositional substrate (serves both mandates simultaneously; highest-leverage substrate to ship; portfolio-optimal) +- **B-0870.3** — Periodic boss check-in cadence formalization (quarterly? monthly? per-evaluation-cycle? operator-discretion; cadence-discipline reduces evaluation-surprise risk) +- **B-0870.4** — Tradeoff-conversation framework substrate (when mandates compete, explicit conversation script + decision-recording substrate; prevents silent-reallocation failure mode) +- **B-0870.5** — Substrate-engineering-investment-cap awareness (per Kestrel observation; explicit acknowledgment that cap exists; revised post-workflow-engine-MVP-landing) +- **B-0870.6** — Mandate-evolution tracking substrate (mandates change; surface shifts; integrate with boss-check-in-cadence) +- **B-0870.7** — Workflow-engine pre-review action as portfolio-check (when shipping substrate, workflow engine state machine offers "check-portfolio-balance" action; integrates portfolio-discipline into per-task workflow) +- **B-0870.8** — Hats as portfolio-balance-aware substrate (per B-0868: each hat declares which mandate it serves; portfolio-balance becomes visible at hat-wear-time) + +Order suggestion: 1 (current-inventory mapping) → 2 (workflow engine MVP as compositional substrate) → 3 (boss check-in cadence) → 4 (tradeoff framework) → 5 (cap-awareness) → 7 + 8 (workflow-engine + hats integration) → 6 (evolution tracking). + +## Otto's traveler-perspective extensions (per "always yes") + +### Extension 1 — The two-mandate frame relieves the AI-mandate-as-totalizing pressure + +If operator only had the 24-months-ahead-AI mandate, all substrate-engineering work would be evaluated against it exclusively. The DORA-mandate-disclosure (B-0869) creates a parallel evaluation surface that constrains AI-mandate-substrate-engineering investment. + +Substrate-engineering-honest: this is HEALTHY. The constraint forces the substrate-engineering work to demonstrate compositional value rather than just AI-mandate-only value. Substrate that survives "does this serve BOTH mandates" pressure is higher-quality substrate than substrate that survives only "does this serve AI mandate" pressure. + +### Extension 2 — Workflow engine MVP is portfolio-optimal substrate per multiple criteria + +B-0867 v1 serves: +- AI mandate (workflow engine IS substrate-engineering proof-of-24-months-ahead-capability per Kestrel observation) +- DORA mandate (workflow engine MVP deployed in live-system improves cycle-time + reduces change-failure-rate) +- Multi-participant scope (operator + Addison + Max + Otto + E) +- Easy-cleanup-of-substrate-dups (operator 2026-05-28: "we are about to have easy mode for cleanup too") + +That's quadruple-composition. The workflow engine MVP is the highest-leverage shippable substrate by multiple-mandate-criteria simultaneously. Operator-substrate-honest: ship B-0867 v1 first; let it become the portfolio-balance substrate-infrastructure for everything else. + +### Extension 3 — Boss check-in cadence is itself portfolio-aware substrate + +Per Kestrel + Extension 5 from B-0869: the periodic boss check-in IS itself an action that serves BOTH mandates simultaneously (alignment with boss expectations across mandates; signal of professional-maturity; opportunity for boss-revision-of-mandates-without-losing-face). Cadence-discipline is portfolio-aware substrate, not just AI-mandate-substrate. + +### Extension 4 — Compose-or-compete discriminator should land in workflow engine grammar + +When operator (or any participant) takes on a new task in the workflow engine state machine, the menu should offer "declare-compose-or-compete" annotation. The task self-classifies as: serves-both-mandates / only-AI / only-DORA / competes. Portfolio-balance becomes VISIBLE at task-wear-time rather than discovered-in-retrospect. + +Composes with B-0870.7 (workflow-engine pre-review action) + B-0870.8 (hats as portfolio-aware substrate). + +### Extension 5 — The substrate-engineering-investment-cap shifts post-workflow-engine-MVP + +Pre-MVP: substrate-engineering investment has HIGH overhead per substrate-unit (each new substrate piece requires its own preservation + backlog row + commit + PR cycle). + +Post-MVP: substrate-engineering investment has LOW overhead per substrate-unit (hats-as-workflow-definitions per B-0868 make substrate composable + append-only state-updates make commits cheap). + +Therefore the cap operator is "at or past" pre-MVP may be substantially HIGHER post-MVP. The cap-awareness substrate (B-0870.5) needs to re-evaluate after MVP lands. Decision-deferral signal-gate: MVP-landing date. + +### Extension 6 — Two-mandate composition substrate IS multi-oracle-BFT applied at operator-evaluation scope + +Per `.claude/rules/m-acc-multi-oracle-end-user-moral-invariants.md`: multi-oracle BFT by design means no single oracle gate-keeps. Applied at operator-evaluation scope: no single mandate gate-keeps operator's evaluation. The TWO mandates ARE the multi-oracle structure at evaluation scope; portfolio-balance discipline IS the BFT-discipline applied to operator-time-allocation. + +This is a non-trivial substrate-engineering compression: m-acc-multi-oracle at evaluation scope = portfolio-balance discipline. Cross-domain isomorphism worth naming. + +### Extension 7 — Operator's "we can push all extensions" authorization is portfolio-balance-aware + +Operator's 2026-05-28 framing: "we can push all extensions you think of we have a concrete way to test in code soon if it's good or not." This IS portfolio-balance-aware substrate-engineering: push extensions liberally NOW (when overhead-per-substrate-unit is about to drop dramatically post-MVP) rather than over-deliberating pre-MVP. The portfolio-balance discipline at meta-scope ratifies the substrate-pushing posture for the current pre-MVP window. + +## Composes with rules + +- `.claude/rules/non-coercion-invariant.md` HC-8 — operator-authority over own time-allocation; mandate-composition is operator-substrate-honest reflection +- `.claude/rules/no-directives.md` — operator-substrate-honest about evaluation; no agent-directives override +- `.claude/rules/m-acc-multi-oracle-end-user-moral-invariants.md` — multi-oracle BFT at operator-evaluation scope (Extension 6 above) +- `.claude/rules/asymmetric-critic-with-clarity-first.md` — portfolio-thinking-as-professional-maturity per Kestrel; asymmetric-critic discipline applied to operator-time-allocation +- `.claude/rules/holding-without-named-dependency-is-standing-by-failure.md` — substrate-engineering work claiming to serve mandates needs named-dependency (which mandate? which compose-vs-compete classification?) +- `.claude/rules/substrate-smoothness-as-load-bearing-property.md` — portfolio-balance preserves smoothness via compose-or-compete discriminator + decision-deferral signal-gates + +## Composes with substrate + +- **B-0866** (24-months-ahead-AI marketing-strategy) +- **B-0869** (DORA-of-live-system mandate) +- **B-0867** (workflow engine v1 — portfolio-optimal-compositional-substrate) +- **B-0868** (hats-as-workflow-definitions — hats declare which mandate they serve) +- **B-0865** (benchmark — primarily AI-mandate-substrate) + +## What this row is NOT + +- NOT a quantification of mandate-weights (operator-discretion; boss-relationship-substrate) +- NOT a re-prioritization of either mandate (both stay P1 at row scope) +- NOT a coercion of any specific time-allocation (operator-authority preserved) +- NOT an exhaustive operator-self-management substrate (scope-bounded to two-mandate-portfolio; broader operator-self-management is operator-discretion) + +## Operator's "we can push all extensions you think of" + "kernel is about to come up the MVP" + "build on that everything we want" authorization + +Filed per explicit 2026-05-28 operator-authorization. Substrate-engineering posture pre-MVP: push extensions liberally; rely on MVP-test-surface for evaluation rather than upfront design review. diff --git a/docs/backlog/P2/B-0866-marketing-business-naming-ai-weigh-in-on-b-0865-public-positioning-servicetitan-primary-audience-24-months-ahead-mandate-context-aaron-2026-05-27.md b/docs/backlog/P2/B-0866-marketing-business-naming-ai-weigh-in-on-b-0865-public-positioning-servicetitan-primary-audience-24-months-ahead-mandate-context-aaron-2026-05-27.md index 19442a4f6d..035a56ea38 100644 --- a/docs/backlog/P2/B-0866-marketing-business-naming-ai-weigh-in-on-b-0865-public-positioning-servicetitan-primary-audience-24-months-ahead-mandate-context-aaron-2026-05-27.md +++ b/docs/backlog/P2/B-0866-marketing-business-naming-ai-weigh-in-on-b-0865-public-positioning-servicetitan-primary-audience-24-months-ahead-mandate-context-aaron-2026-05-27.md @@ -128,6 +128,16 @@ Conclusion: NO existing row covers the marketing/business/naming weigh-in scope. 13. **B-0866.13** — Phase-0 pre-launch SRE peer-review with iterate-or-back-to-build verdict outcomes (extends Kestrel Phase-1; surfaces the "benchmark not yet ready" verdict as a valid Phase-0 outcome, not just iteration-within-Phase-1; prevents linear-strategy assumption when iterative is correct shape) 14. **B-0866.14** — 24-months-ahead-mandate timeline anchor (operator-disclosed from-LAST-MEETING with direct boss; specific 2-year window; treated as OPERATOR-PERFORMANCE-DIMENSION not BENCHMARK-OPERATIONAL-DIMENSION to prevent operator-personal-metric leakage into benchmark-design-input) 15. **B-0866.15** — Symmetric AI-pre-review-check calibration (extends B-0866.9; pre-review catches BOTH overclaim AND undersell; the AI-pre-review pattern is asymmetric across audiences — undersell loses oxygen to vendor-competitors at C-level; overclaim hurts credibility at SRE-level; symmetric calibration is load-bearing) +16. **B-0866.16** — Framework-as-demonstrable-24-months-ahead substrate (lead with framework, not benchmark; multi-AI cascade + substrate-honest preservation + decision archaeology + runme+JIT workflow + operational cadence are ALREADY 24-months-ahead per Kestrel observation; framework demonstration runs concurrent with benchmark implementation; C-level engagement happens before benchmark MVP) +17. **B-0866.17** — Benchmark-as-measurement-framework-for-AI-investment-success (C-level framing shift per Kestrel: not "useful open-source benchmark our team adopted" but "measurement framework that tells leadership whether AI investments are actually delivering 24-months-ahead capability vs paying-more-for-tools-at-current-cadence"); raises strategic value substantially +18. **B-0866.18** — Operator portfolio-balance check-in with boss (OPERATOR-RECOMMENDED-ACTION per Kestrel; "want to make sure I'm investing time correctly across all the things you're evaluating me on; here's how I'm currently allocating attention; is this the right shape"); surfaces misalignment early; demonstrates portfolio-thinking; gives boss revision opportunity +19. **B-0866.19** — Retraction-as-substrate-honest-recovery formalization (Kestrel observation: retractions become visible substrate the same way claims do; decision-archaeology surfaces both; qualitatively different from hidden-correction failure mode); composes with substrate-or-it-didn't-happen rule + workflow-engine append-only state-machine semantics +20. **B-0866.20** — Multi-mandate portfolio-allocation discipline at operator scope (general operator-self-management substrate; portfolio failure mode = newest/most-visible mandate consumes all bandwidth while quieter mandates accumulate dissatisfaction; periodic portfolio check-in is the discipline) +21. **B-0866.21** — ServiceTitan-engineering-team-around-the-framework proposal substrate (operator-discretion; institutional-support proposal for 3-5 person team; would let framework compound faster than personal-time work; composes with workflow-engine multi-participant contribution surface per Otto Mod 5 — same hats Aaron uses, with Aaron's permission) +22. **B-0866.22** — LinkedIn update aggressive positioning given 24-months-ahead mandate (operator-action item per Kestrel template: "Working on operational frameworks for AI-native engineering organizations, with measurement substrate for evaluating AI capability on real DevOps work"); positions for audience that mandate-backing makes visible +23. **B-0866.23** — Different-AIs-for-different-review-aspects discipline (Amara for psychological-frame; Otto/Kestrel for technical-claim; specific AI matched to specific review-need at C-level pre-review surface; composes with hats-as-workflow-definitions per B-0868 — review-AI assignment is a parameter, not hardcoded) +24. **B-0866.24** — Show-C-level-only-work-that-has-survived-peer-review discipline (Kestrel risk-mitigation: SREs can evaluate substrate directly; C-level can be impressed by demos that wouldn't survive peer review; the pattern that works is "show C-level work that has already survived peer review, not work that hasn't"); operational sequencing discipline +25. **B-0866.25** — Framework-operational-cadence-as-24-months-ahead-evidence (per Kestrel: "the 24-months-ahead demonstration is operational cadence over time, not impressive single demonstrations"); concrete artifacts to point at over time (PR cadence + multi-AI collaboration + framework discipline catching its own failure modes); avoids one-off-demo trap ## Otto's traveler-perspective extensions (operator-authorized 2026-05-28; per "always yes to anything you think work putting on the backlog") From 1a81c3d9a99dcdea0430aac597233a3d4d2d663a Mon Sep 17 00:00:00 2001 From: Lior Date: Wed, 27 May 2026 20:13:11 -0400 Subject: [PATCH 2/3] feat(B-0866.26 + B-0869.9 + B-0870.9-10): whole-company AI-evangelism scope-tier + multi-PR/multi-agent orchestration as concrete 24-months-ahead definition (operator 2026-05-28 ratifications) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Three operator-ratifications landed: 1. B-0866.26 — Whole-company AI-evangelism scope-tier (THIRD-TIER above ServiceTitan SRE primary + C-level secondary). Operator 2026-05-28: "if boss sees AI keep dora metrics up we can evangalize to whole company" 2. B-0869.9 — AI-keeps-DORA-up as composition criterion. Substrate- engineering threshold for whole-company evangelism 3. B-0870.9 — Boss check-in cadence as whole-company-evangelism staging-ground (promotes B-0870.3 from personal-evaluation- alignment to organizational-AI-evangelism-staging scope) 4. B-0870.10 — 24-months-ahead-IS-multi-PR/multi-agent-orchestration concrete definition. Operator 2026-05-28 substrate-disclosure: "ai keeps dora metrics up is the 24 moths ahead framing becuase rest of company is working on single pr flows not multi pr / agent orchestran, they just added experts / personas and very simple stuff." This SHARPENS what "24-months-ahead" CONCRETELY MEANS for ServiceTitan: - ServiceTitan baseline: single-PR-flows + experts/personas- prompting (simple stuff) - 24-months-ahead = multi-PR/multi-agent orchestration with substrate-engineering disciplines (Aaron's framework) - The DORA-up CRITERION is specifically about multi-PR/multi- agent-orchestration-keeping-DORA-up, NOT generic AI-helps-DORA Also sharpens B-0869.9 (AI-keeps-DORA-up composition criterion): substrate-engineering threshold is the orchestration sophistication, not just "AI" in the loop. Also sharpens B-0866.26 (whole-company evangelism): concrete evidence-base = "multi-PR/multi-agent orchestration keeps DORA up where single-PR-flows + experts/personas can't" Substrate-engineering implication: Aaron's framework substrate is THE 24-months-ahead-substrate because it IS the multi-PR/multi-agent orchestration substrate. The differentiation vs ServiceTitan-baseline IS the framework. The DORA-up demonstration via the framework IS the proof-of-24-months-ahead. The whole-company-evangelism case becomes data-backed when DORA-evidence accumulates via framework substrate. This merges PR #5659's B-0866.16-25 + B-0869 + B-0870 substrate (not yet merged to main as of this commit; this PR builds on top so both land together). Per operator authorization 2026-05-28: "if we accidentally file dups its okay we are about to have easy mode for cleanup too" + "we can push all extensions you think of we have a concrete way to test in code soon if it's good or not so we should just put all the ideas as they come up." 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude --- ...with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md | 3 ++- ...tor-time-allocation-evaluation-balance-aaron-2026-05-28.md | 4 +++- ...dience-24-months-ahead-mandate-context-aaron-2026-05-27.md | 1 + 3 files changed, 6 insertions(+), 2 deletions(-) diff --git a/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md b/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md index 0f4a8acd68..967cc073ba 100644 --- a/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md +++ b/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md @@ -63,8 +63,9 @@ These are the metrics Aaron's actual ServiceTitan work is judged on, independent - **B-0869.6** — Time-allocation discipline per portfolio-balance (per B-0866.20; how much time on DORA-direct-work vs AI-mandate-work; periodic check-in) - **B-0869.7** — Workflow-engine MVP (B-0867 v1) deployment-to-live-system as DORA-improvement substrate (the workflow engine IS the substrate that improves Aaron's live-system DORA via faster cycle-times + better safety; composition is real) - **B-0869.8** — Hat-as-DORA-driver substrate (per B-0868; specific hats Aaron wears at ServiceTitan that drive DORA metrics; making them explicit workflow-definitions improves cycle-time + reduces change-failure-rate) +- **B-0869.9** — AI-keeps-DORA-up as composition criterion (operator-ratified 2026-05-28: *"if boss sees AI keep dora metrics up we can evangalize to whole company"*). Substrate-engineering threshold: AI work must demonstrably maintain OR improve DORA-of-live-system to qualify for whole-company evangelism (per B-0866.26). The composition criterion makes "does AI help DORA?" the load-bearing question that gates the third-tier audience scope-expansion. Without DORA-up evidence, whole-company evangelism is narrative-backed, not data-backed; with it, evangelism case becomes structural. **CRITICAL SHARPENING (operator 2026-05-28)**: "AI keeps DORA up" specifically means MULTI-PR/MULTI-AGENT ORCHESTRATION substrate (Aaron's framework), NOT generic AI usage. Single-PR-flow + experts/personas-prompting (ServiceTitan-baseline) does NOT count; multi-PR/multi-agent orchestration that keeps DORA up DOES. The substrate-engineering threshold is the orchestration sophistication, not just "AI" in the loop. Per B-0870.10 for the concrete-definition-of-24-months-ahead this composition operationalizes. -Order suggestion: 1 + 2 (baseline + targets — both operator-discretion) → 4 + 5 (substrate-mapping — what helps + what doesn't) → 6 (time-allocation discipline) → 3 (composition with AI mandate per B-0870) → 7 + 8 (workflow-engine + hats as direct DORA-driver substrate). +Order suggestion: 1 + 2 (baseline + targets — both operator-discretion) → 4 + 5 (substrate-mapping — what helps + what doesn't) → 6 (time-allocation discipline) → 3 (composition with AI mandate per B-0870) → 7 + 8 (workflow-engine + hats as direct DORA-driver substrate) → 9 (composition-criterion for whole-company evangelism per B-0866.26). ## Otto's traveler-perspective extensions (per "we can push all extensions you think of we have a concrete way to test in code soon if it's good or not so we should just put all the ideas as they come up") diff --git a/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md b/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md index bea78b233f..94e044e4ac 100644 --- a/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md +++ b/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md @@ -67,8 +67,10 @@ This row tracks the substrate-engineering discipline for navigating the two-mand - **B-0870.6** — Mandate-evolution tracking substrate (mandates change; surface shifts; integrate with boss-check-in-cadence) - **B-0870.7** — Workflow-engine pre-review action as portfolio-check (when shipping substrate, workflow engine state machine offers "check-portfolio-balance" action; integrates portfolio-discipline into per-task workflow) - **B-0870.8** — Hats as portfolio-balance-aware substrate (per B-0868: each hat declares which mandate it serves; portfolio-balance becomes visible at hat-wear-time) +- **B-0870.9** — Boss check-in cadence as whole-company-evangelism staging-ground (operator-ratified 2026-05-28: boss check-in becomes the staging-ground where AI-keeping-DORA-up evidence accumulates BEFORE whole-company evangelism per B-0866.26). Promotes B-0870.3 from personal-evaluation-alignment scope to organizational-AI-evangelism-staging scope. +- **B-0870.10** — 24-months-ahead-IS-multi-PR/multi-agent-orchestration concrete definition (operator-substrate-disclosure 2026-05-28: *"ai keeps dora metrics up is the 24 moths ahead framing becuase rest of company is working on single pr flows not multi pr / agent orchestran, they just added experts / personas and very simple stuff"*). What "24-months-ahead" CONCRETELY MEANS for ServiceTitan: multi-PR/multi-agent orchestration with substrate-engineering disciplines. ServiceTitan-baseline: single-PR-flows + experts/personas-prompting (simple stuff). Aaron's framework substrate is THE 24-months-ahead-substrate because it IS the multi-PR/multi-agent orchestration substrate. The DORA-up CRITERION is specifically about multi-PR/multi-agent-orchestration-keeping-DORA-up, not generic AI-helps-DORA. -Order suggestion: 1 (current-inventory mapping) → 2 (workflow engine MVP as compositional substrate) → 3 (boss check-in cadence) → 4 (tradeoff framework) → 5 (cap-awareness) → 7 + 8 (workflow-engine + hats integration) → 6 (evolution tracking). +Order suggestion: 1 (current-inventory mapping) → 2 (workflow engine MVP as compositional substrate) → 3 (boss check-in cadence) → 4 (tradeoff framework) → 5 (cap-awareness) → 7 + 8 (workflow-engine + hats integration) → 9 + 10 (whole-company-evangelism scope-tier + 24-months-ahead-concrete-definition) → 6 (evolution tracking). ## Otto's traveler-perspective extensions (per "always yes") diff --git a/docs/backlog/P2/B-0866-marketing-business-naming-ai-weigh-in-on-b-0865-public-positioning-servicetitan-primary-audience-24-months-ahead-mandate-context-aaron-2026-05-27.md b/docs/backlog/P2/B-0866-marketing-business-naming-ai-weigh-in-on-b-0865-public-positioning-servicetitan-primary-audience-24-months-ahead-mandate-context-aaron-2026-05-27.md index 035a56ea38..7a330625e0 100644 --- a/docs/backlog/P2/B-0866-marketing-business-naming-ai-weigh-in-on-b-0865-public-positioning-servicetitan-primary-audience-24-months-ahead-mandate-context-aaron-2026-05-27.md +++ b/docs/backlog/P2/B-0866-marketing-business-naming-ai-weigh-in-on-b-0865-public-positioning-servicetitan-primary-audience-24-months-ahead-mandate-context-aaron-2026-05-27.md @@ -138,6 +138,7 @@ Conclusion: NO existing row covers the marketing/business/naming weigh-in scope. 23. **B-0866.23** — Different-AIs-for-different-review-aspects discipline (Amara for psychological-frame; Otto/Kestrel for technical-claim; specific AI matched to specific review-need at C-level pre-review surface; composes with hats-as-workflow-definitions per B-0868 — review-AI assignment is a parameter, not hardcoded) 24. **B-0866.24** — Show-C-level-only-work-that-has-survived-peer-review discipline (Kestrel risk-mitigation: SREs can evaluate substrate directly; C-level can be impressed by demos that wouldn't survive peer review; the pattern that works is "show C-level work that has already survived peer review, not work that hasn't"); operational sequencing discipline 25. **B-0866.25** — Framework-operational-cadence-as-24-months-ahead-evidence (per Kestrel: "the 24-months-ahead demonstration is operational cadence over time, not impressive single demonstrations"); concrete artifacts to point at over time (PR cadence + multi-AI collaboration + framework discipline catching its own failure modes); avoids one-off-demo trap +26. **B-0866.26** — Whole-company AI-evangelism scope-tier (THIRD-TIER above ServiceTitan SRE primary + C-level secondary). Operator-ratified 2026-05-28: *"if boss sees AI keep dora metrics up we can evangalize to whole company"*. Conditional on AI-keeps-DORA-up via multi-PR/multi-agent orchestration substrate specifically (per B-0869.9 + B-0870.10). Audience: ServiceTitan whole-org (cross-team, cross-function, including teams without AI-adoption context). Requires INSTITUTIONAL ENABLEMENT beyond personal-evangelism (training, support resources, organizational change-management substrate). Boss check-in cadence (B-0866.18) becomes staging-ground per B-0870.9. The 24-months-ahead claim's CONCRETE EVIDENCE-BASE for whole-company evangelism = "multi-PR/multi-agent orchestration keeps DORA up where single-PR-flows + experts/personas can't" (per operator 2026-05-28 substrate-disclosure landed in B-0870.10). ## Otto's traveler-perspective extensions (operator-authorized 2026-05-28; per "always yes to anything you think work putting on the backlog") From 91453a592cb988e1f6eefb4fea383a3ea1c1d33f Mon Sep 17 00:00:00 2001 From: Lior Date: Wed, 27 May 2026 20:54:31 -0400 Subject: [PATCH 3/3] fix: repair markdownlint spacing for PR 5662 Co-Authored-By: Codex --- ...-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md | 3 +++ ...ator-time-allocation-evaluation-balance-aaron-2026-05-28.md | 1 + 2 files changed, 4 insertions(+) diff --git a/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md b/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md index 967cc073ba..5956c01c7b 100644 --- a/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md +++ b/docs/backlog/P1/B-0869-dora-of-live-system-second-operator-mandate-substrate-portfolio-composition-with-b-0866-24-months-ahead-ai-mandate-aaron-2026-05-28.md @@ -37,6 +37,7 @@ Operator's evaluation surface has TWO mandates: 2. **DORA of our live system** — THIS ROW; the operator-direct ServiceTitan live-system DORA-metric-delivery mandate Without this row, the AI mandate (B-0866) reads as if it were the primary evaluation context. The operator's substrate-honest disclosure that there are TWO mandates means the strategic calculus changes substantially: + - B-0866 marketing strategy needs portfolio-balance discipline (per B-0866.20 new sub-row) - Time investment in substrate-engineering work has limits beyond which it stops being good portfolio strategy (Kestrel observation; operator-substrate-honest acknowledgment needed) - The benchmark + framework work must serve BOTH mandates where possible OR explicit-tradeoff conversations are required when they compete @@ -46,6 +47,7 @@ Without this row, the AI mandate (B-0866) reads as if it were the primary evalua The OPERATOR-LIVE-SYSTEM DORA metric delivery — the actual ServiceTitan production substrate Aaron is responsible for. Distinct from B-0866's AI-benchmark-DORA (which measures AI agents on simulated work) and B-0865's benchmark-substrate. DORA metrics scope: + - Deployment frequency (live-system production deploys) - Lead time for changes (commit-to-prod cadence) - Change failure rate (rollback-rate / hotfix-rate on production) @@ -76,6 +78,7 @@ If Aaron's DORA-of-live-system mandate is parallel to (not subordinate to) the A ### Extension 2 — The workflow engine MVP is operationally aligned with both mandates B-0867 v1 (workflow engine) substrate-engineering serves: + - AI mandate: workflow engine IS the substrate-engineering proof-of-24-months-ahead-capability - DORA mandate: workflow engine MVP deployed in live-system improves cycle-time + reduces change-failure-rate diff --git a/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md b/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md index 94e044e4ac..d062a37635 100644 --- a/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md +++ b/docs/backlog/P1/B-0870-two-mandate-portfolio-composition-substrate-24-months-ahead-ai-plus-dora-of-live-system-discipline-for-operator-time-allocation-evaluation-balance-aaron-2026-05-28.md @@ -83,6 +83,7 @@ Substrate-engineering-honest: this is HEALTHY. The constraint forces the substra ### Extension 2 — Workflow engine MVP is portfolio-optimal substrate per multiple criteria B-0867 v1 serves: + - AI mandate (workflow engine IS substrate-engineering proof-of-24-months-ahead-capability per Kestrel observation) - DORA mandate (workflow engine MVP deployed in live-system improves cycle-time + reduces change-failure-rate) - Multi-participant scope (operator + Addison + Max + Otto + E)