01
Prompt Library Architect
Turns scattered prompts and notes into a versioned, searchable library without silently merging conflicts.
taxonomyversioningdedupe
AI OPERATIONS · 10 SYSTEM PACK · MULTILINGUAL
Build the operating layer around AI.
Ten focused systems for the work around the model: context, verification, design handoff, multi-agent coordination, evidence, UI specification, failure recovery, distribution and evaluation.
Get all 10 · $39 →One-time purchase10 independent systemsEnglish-firstMultilingual outputSingle-user license
WHY THIS PACK
Less prompt collecting.
More operational control.
Each system solves one recurring failure point. Use only the system that matches the job, replace the bracketed fields with real context, and choose your output language.
01 · ContextGive AI the right state, not the whole archive.
02 · ExecuteRoute complex work into finite operations.
03 · VerifySeparate evidence from confident-looking output.
04 · ImproveEvaluate variants instead of trusting one demo.
THE 10 SYSTEMS
A toolbox for AI operations.
02
AI Context Pack Builder
Compresses project state into a small source-of-truth packet with evidence, constraints and handoff instructions.
contexthandoffsource of truth
03
AI Output QA Auditor
Audits claims, calculations, links and assumptions; returns PASS, PASS WITH FIXES or FAIL.
hallucinationQAevidence
04
Design → Code Handoff Architect
Converts visual intent into named UI patterns, component states, responsive rules and acceptance tests.
Claude CodeCodexUI spec
05
Multi-Agent Workflow Orchestrator
Defines ownership, parallel work, handoffs, shared state, loop guards and final integration.
agentsroutingstop conditions
06
Evidence-to-Decision Synthesizer
Separates fact, estimate, inference and preference before comparing options and recommending action.
researchdecisionuncertainty
07
UI Pattern Brief Builder
Replaces “make it modern” with concrete navigation, form, feedback, commerce and command-surface patterns.
designmobilemotion
08
Automation Recovery Planner
Designs bounded retries, idempotency, fallback, monitoring and human escalation before production failure.
recoveryretriesmonitoring
09
Distribution Repurposing Router
Turns one source asset into platform-native value posts with disclosure, UTMs and anti-spam safeguards.
distributionCTAmeasurement
10
Prompt Evaluation Lab
Builds representative test cases, rubrics and winner rules so prompt changes are evaluated rather than admired.
evalsA/Bfailure taxonomy
FULL PACK
10 systems.
One operating layer.
Use with ChatGPT, Claude and other instruction-following AI tools. Set [OUTPUT_LANGUAGE] to Japanese, English or another language you need.
No guaranteed revenue, productivity or accuracy outcome. Verify external facts, permissions and high-impact actions. Single-user license; no redistribution, resale or public mirroring.