brewtoolstext-optimizelisted

# Text & File Optimizer ## Step 0: Load Rules > **REQUIRED:** Read `references/rules-review.md` before ANY optimization. > If file not found -> ERROR + STOP. Do not proceed without rules reference. ## Modes Parse `$ARGUMENTS`: `-l`/`--light` | `-s`/`--standard` | `-d`/`--deep` | `-x`/`--max` | no flag -> medium (default) or auto-detect. | Mode | Flag | Target | Compression | Human-readable | Verification | |------|------|--------|-------------|----------------|--------------| | Light | `-l`, `--light` | Any | Minimal | Yes | None | | Medium | _(default)_ | Any | Moderate | Yes | Self-check (fact inventory) | | Standard | `-s`, `--standard` | Docs, README | 30-50% | Yes | 1 round (>=98%) | | Deep | `-d`, `--deep` | CLAUDE.md, system prompts, agent/skill defs, KNOWLEDGE | 2-3x | No (LLM-only) | 1-2 rounds (>=95%) | | Max | `-x`, `--max` | CLAUDE.md, system prompts, KNOWLEDGE | 3-4x | No (LLM-only) | 2 mandatory (>=95% + 100% sub-gate) | ## Loss Budget per Mode Content essence is untouchable at light/medium/standard; small deliberate loss is allowed only at deep/max — explicitly reported. Dedup-merged facts count as preserved, never as loss. | Mode | Semantic match target | Allowed loss | |------|----------------------|--------------| | Light | 100% | None — wording cleanup only | | Medium | 100% | None — restructure, zero fact loss (self-check) | | Standard | >= 98% | None intended; verification patches any slip | | Deep | >= 95% | Word-level drops (A.2, ledgered, gate-