dotclaude

Metric, SQL, dbt, dashboard, semantic-layer, and BI validation discipline for high-stakes analytics work. Use for grain, joins, nulls, filters, freshness, lineage, reconciliation, dashboard semantics, and decision relevance checks.

Compact task-state and handoff ledger for long Opus 4.8 High sessions, subagent orchestration, compaction, and resumable workflows. Use to preserve decisions, files, commands, evidence, risks, and next actions while dropping noisy logs.

Classify high-rigor work into S0-S4 scope levels and choose the minimal sufficient subagent set. Use for ultracode-high orchestration, complex coding, analytics, BI, strategy, writing, audits, migrations, and multi-domain tasks.

Evidence, claim, assumption, and uncertainty ledger for high-rigor agent work. Use when synthesizing research, analytics, code review, strategy, or writing so important claims are labeled as verified, calculated, inferred, assumed, or unknown with source evidence.

Final response contract for ultracode-high and high-rigor subagent workflows. Use to synthesize work into a relayable result with changes, verification, final-gate improvements, risks, and next actions.

Gate vocabulary and runner support for ultracode-high. Use to define precondition and completion gates, prefer deterministic command/file/grep gates where possible, flag attestation gates honestly, and run gates with the bundled gate_check.py helper.

Fork-style final gate for code and technical deliverables. Use after the draft result exists to check completeness, constraints, verification, risks, regression exposure, and final answer quality before handoff.

Fork-style read-only repository exploration workflow for Opus 4.8 High. Use when a task needs isolated codebase research, architecture mapping, dependency tracing, or risk discovery without editing files.

Fork-style adversarial review workflow for the current diff or a proposed code change. Use to find correctness, regression, security, performance, maintainability, test, and documentation issues before final delivery.

Fork-style verification workflow for running or planning lint, typecheck, tests, builds, dbt checks, and other project validation commands. Use to return compact verification cards without polluting the main context with logs.

Coding implementation contract for Opus 4.8 High workers. Use when making scoped code changes, refactors, migrations, or docs updates that must follow repo conventions, minimize diff size, preserve behavior, add tests when appropriate, and report verification honestly.

Performance review discipline for algorithms, database queries, rendering, network behavior, caching, concurrency, memory, benchmarks, and scalability risks in code, data, and frontend work.

Security review discipline for auth, authorization, injection, secrets, permissions, dependency, supply-chain, privacy, logging, and data exposure risks in high-rigor code and architecture work.

Known-answer self-test harness for ultracode-high. Use to verify that the orchestration mechanism fires: plan critique blocks bad manifests, verification catches planted failures, constraints survive long runs, unknowns are flagged, and enumerable tasks close completely.

Strategy and business decision rigor for product, GTM, pricing, operations, market, roadmap, and executive recommendations. Use to evaluate options, criteria, evidence, economics, risks, assumptions, execution, measurement, and decision quality.

Build a compact, explicit intake contract for messy requests before planning or delegation. Use inside ultracode-high and other high-rigor workflows to identify the goal, artifact, constraints, acceptance criteria, missing decisions, risk level, and verification target.

Opus 4.8 High-only orchestration mode for complex coding, analytics, BI, strategy, and writing tasks. Use when work spans multiple files or domains, carries material risk, needs verification, or benefits from subagent fan-out, adversarial review, and a final integration gate. Keeps every planned skill and subagent on claude-opus-4-8 with effort high.

Enterprise dashboard and operational UI review discipline for information hierarchy, first-screen answers, accessibility, scanning, filters, drill paths, annotations, alerts, and misread risks.

Standard verification gates for code, analytics, BI, writing, strategy, and documentation deliverables. Use before final delivery and inside review or test-runner subagents to record passed, failed, skipped, and not-applicable checks with evidence.

High-rigor writing and editing discipline for PRDs, memos, docs, executive narratives, release notes, and plans. Use for audience fit, thesis, structure, evidence, terminology, concision, voice, and actionable next steps.

Makes Claude Opus 4.8 behave with GPT-5.5-Pro-like operating discipline for analytics, business intelligence, data science, finance, and strategy work: deeper planning, persistent tool use, evidence-first reasoning, artifact completion, and rigorous self-verification.

Operating mode that makes ONE Claude Opus 4.8 context reproduce the rigor of Opus 4.8 in multi-agent "ultracode" mode for the DOING of analytics, BI, and strategy work — the analysis itself, not the spec for it. Use when the cost of a wrong result is more than trivial AND a stakeholder will act on it: metric pulls and reconciliations, "why did X move" investigations, experiment readouts, trend/funnel/segmentation analysis, forecasts, KPI definitions, dashboard and report numbers, and quantitatively load-bearing market, prioritization, and executive recommendations. Literal triggers: "ultracode" (unambiguous — always fires this skill), "leadership will see this", "high visibility", "double-check the numbers", "why did X move/drop/spike". Do NOT trigger for exploratory or throwaway queries no one will act on. Routing vs siblings: not a pre-implementation spec (use analytics-spec-builder). Overlaps gpt-5.5-xhigh, the other analytics/BI/strategy operating mode — prefer THIS skill when the task is a single high-st

Build, review, refactor, and modernize Expo and React Native apps for iOS, Android, and web. Use for Expo Router, app architecture, beautiful mobile UI, design systems, animations, accessibility, performance, native modules, testing, EAS Build/Update, and improving existing Expo code.

Rigorous second-pass self-review and improvement skill for analytics, BI, strategy, code, and writing deliverables. Use when the user says "you can and MUST do better", asks for critique, review, audit, refine, improve, revise, final check, or wants a higher-quality answer before delivery.

Use this skill for analytics, business intelligence, data modeling, dashboarding, financial analysis, strategic planning, executive decision support, large document/data synthesis, or any ambiguous high-stakes analytical task. It makes Claude Opus 4.8 behave in a GPT-5.5-xhigh-inspired operating mode: outcome-first, tool-grounded, verification-heavy, concise, strategically complete, and explicit about evidence, assumptions, risks, and decision implications.

Use this skill proactively when the user asks Claude Code to create, tighten, review, or improve a specification for a scoped analytics task/project—not a whole software product—including dashboards, reports, SQL/dbt models, metric definitions, semantic-layer changes, experiment/cohort analyses, data quality checks, reconciliations, KPI automation, notebooks, or analytics refactors. Ask targeted user questions, inspect relevant repo context, convert ambiguity into explicit assumptions/decisions, and produce an implementation-ready spec with scope, data sources, metric logic, grain, filters, validation plan, acceptance criteria, risks, and open questions before coding.