| name | source-command-high-taste-writing-os |
| description | Run high-taste-writing-os when the user wants V4-level remarkable output, high-taste generation, output elevation, enrichment, audit, practitioner-grade artifacts, publish-ready/client-ready writing, or work that feels flat, consultant-clean, generic, AI-shaped, or structurally correct but not compelling. |
source-command-high-taste-writing-os
Use this skill when the user asks for /high-taste-writing-os, V4-level
output, remarkable output, publish-ready high-taste work, client-ready strategy,
or an elevation/enrichment pass on a flat or consultant-clean artifact.
Operator Core Alignment
This project wrapper follows .agent/workflows/high-taste-writing-os.md as the
canonical behavior source. It must stay a thin compatibility wrapper and
preserve:
- high-taste as an operating wrapper that selects the right owner, not a persona
- V4 golden sample and repeatability packet loaded before substantial output
- Context Load Packet before drafting
- one functional owner and bounded support lanes
- source spine and Claim/Proof Ledger when factual
- Human Wound/Stakes before composing
- Taste Evidence Ledger for high-stakes generation or elevation
- Orchestration Receipt before delivery
- no publishing, outreach, connector writes, paid tools, destructive actions, or real Codex subagents without explicit approval
Command Template
Read and execute the workflow at .agent/workflows/high-taste-writing-os.md.
