winning-writinglisted

# Winning Writing — mode-aware Source: Rachel Konrad's GSB guest lectures, Heidi Roizen's rules, 25 model letters, live student critiques. Rule library is in `rules/catalog.json`; mode definitions in `rules/modes.json`. Full playbook in `references/cold_outreach_playbook.md`. ## Invocation ``` /winning-writing <mode> "paste draft here" ``` Where `<mode>` is one of: `cold-email`, `memo`, `essay`, `profile`, `readme`, `linkedin-post`, or `default`. If no mode is given, defaults to `default` (every rule in the catalog fires). The mode arg filters the catalog to rules tagged with that mode, applies any per-mode severity overrides from `modes.json`, and applies the mode's target-length check if one is set. ## What this skill does on every invocation 1. **Read `rules/modes.json`** to get the active mode's definition (label, description, target length, severity overrides). 2. **Read `rules/catalog.json`** and filter to rules whose `modes` array includes the active mode (or the active mode is `default`). 3. **Apply per-mode severity overrides** to the filtered rules. 4. **Critique the draft** against the filtered list. For each rule, name the rule, quote the failing span, give the suggested rewrite, and cite the rule id so the user can look it up. 5. **Apply the length check** if the active mode declares a `target_words` value. 6. **Return** a structured critique with critical issues first, then issues, then warns. End with a clean rewrite of the draft. ## Output format (unch