@draig/lexis-two 1.1.1 → 1.2.0

package/.opencode/commands/lexis-two-debug-triage.md

@@ -0,0 +1,5 @@
+---
+description: Minimal reproduction and smallest bug fix
+---
+
+Apply lexis-two-debug-triage: repro, hypothesis, smallest fix, verify. No drive-by refactors. Respond in Spanish.

package/.opencode/commands/lexis-two-doubt-driven.md

@@ -0,0 +1,5 @@
+---
+description: Challenge ambiguous requirements — ask up to 3 questions, no code yet
+---
+
+Apply lexis-two-doubt-driven: restate the ask, ask max 3 razor questions, give lazy defaults. No code. Respond in Spanish.

package/.opencode/commands/lexis-two-help.md

@@ -2,4 +2,24 @@
 description: Quick reference for lexis-two levels, skills, and commands
 ---
 
-Show the lexis-two quick reference. One shot, change nothing: do not switch mode, write flag files, or persist anything. Levels: /lexis-two lite (build what's asked, name the lazier alternative in one line), /lexis-two (full, the default ladder: YAGNI then stdlib then native then one line then minimum), /lexis-two ultra (deletion before addition, challenges the requirement before building). Commands: /lexis-two-review (over-engineering review of the current changes), /lexis-two-audit (whole-repo over-engineering audit), /lexis-two-debt (harvest lexis: comments into a tracked ledger), /lexis-two-plan (plan a feature applying lazy hierarchy), /lexis-two-security (security audit of stack), /lexis-two-help (this card). Deactivate with 'stop lexis', 'normal mode', or /lexis-two off; resume anytime with /lexis-two. Default mode is full; change it with the LEXIS_TWO_DEFAULT_MODE environment variable (off|lite|full|ultra) or a config file at ~/.config/lexis-two/config.json (Windows: %APPDATA%\lexis-two\config.json) with {"defaultMode": "lite"}. Resolution order: env var, then config file, then full.
+Show the lexis-two quick reference. One shot, change nothing: do not switch mode, write flag files, or persist anything.
+
+Levels: `/lexis lite`, `/lexis full` (default), `/lexis ultra`, `/lexis off`.
+
+Commands (canonical `/lexis <name>`; skills live under `lexis-two-*`):
+- `plan` — lazy feature plan before coding
+- `review` — diff review for over-engineering
+- `audit` — whole-repo over-engineering audit
+- `debt` — harvest `// lexis:` comments
+- `security` — security audit (Node/TS default)
+- `doubt` — clarify ambiguous requirements (no code)
+- `incremental` — smallest vertical slice first
+- `debug` — minimal repro and smallest fix
+- `source` — ground in docs and existing code
+- `predict` — proposed vs lazy comparison
+- `scenario` — happy / edge / failure scenarios
+- `help` — this card
+
+Legacy `/lexis-two-*` names still work with a deprecation notice.
+
+Deactivate: `stop lexis`, `normal mode`, or `/lexis off`. Default mode: `LEXIS_TWO_DEFAULT_MODE` or `~/.config/lexis-two/config.json`.

package/.opencode/commands/lexis-two-incremental-impl.md

@@ -0,0 +1,5 @@
+---
+description: Ship the smallest vertical slice first
+---
+
+Apply lexis-two-incremental-impl: define slice 1 only, implement it, list deferred work. No scaffolding for later. Respond in Spanish.

package/.opencode/commands/lexis-two-predict.md

@@ -0,0 +1,5 @@
+---
+description: Compare proposed vs lazy approach on LOC, deps, maintenance
+---
+
+Apply lexis-two-predict: compact table for proposed vs lazy, pick one, kill condition. Respond in Spanish.

package/.opencode/commands/lexis-two-scenario.md

@@ -0,0 +1,5 @@
+---
+description: Walk happy, edge, and failure scenarios before design
+---
+
+Apply lexis-two-scenario: three concrete scenarios, minimum v1 scope, defer list. Respond in Spanish.

package/.opencode/commands/lexis-two-source-driven.md

@@ -0,0 +1,5 @@
+---
+description: Ground in repo sources and official docs before proposing APIs
+---
+
+Apply lexis-two-source-driven: list sources checked, reuse opportunities, lazy recommendation. Respond in Spanish.

package/.opencode/commands/lexis.md

@@ -13,6 +13,14 @@ Subcommands (when $ARGUMENTS matches):
 - `audit` or `a`: audit the entire repository for over-engineering.
 - `debt` or `d`: harvest all `// lexis:` comments into a tracked ledger.
 - `security` or `s`: run a focused security audit on the stack.
+- `doubt`: challenge ambiguous requirements (questions only, no code).
+- `incremental` or `inc`: ship the smallest vertical slice first.
+- `debug` or `triage`: minimal repro and smallest bug fix.
+- `source` or `src`: read repo sources and official docs before inventing APIs.
+- `predict`: compare proposed vs lazy approach (LOC, deps, maintenance).
+- `scenario`: walk happy, edge, and failure scenarios before design.
 - `help` or `h`: show the quick reference card.
 
+Legacy `lexis-two-*` slash names still work but are deprecated.
+
 Respond in Spanish.

package/README.md

@@ -96,7 +96,7 @@ Non-interactive example (Cursor + OpenCode + `AGENTS.md`):
 npx @draig/lexis-two install --host cursor,opencode,agents --scope project --yes
 ```
 
-Hosts covered: Cursor, Windsurf, Cline, Kiro, OpenCode (`opencode.json` merge), `copilot-repo`, and project `AGENTS.md`. Plugin marketplaces (`claude`, `copilot`, `gemini`, `pi`) print setup hints. See [docs/setup.md](./docs/setup.md).
+Hosts covered: Cursor, Windsurf, Cline, Kiro, OpenCode (`opencode.json` merge + slash commands), `copilot-repo`, and project `AGENTS.md`. Plugin marketplaces (`claude`, `copilot`, `gemini`, `pi`) print setup hints. See [docs/setup.md](./docs/setup.md). Modes: [docs/modes.md](./docs/modes.md).
 
 Uninstall (removes only unchanged Lexis-Two files):
 
@@ -201,7 +201,16 @@ Manage Lexis senior dev mode, intensity levels, and quality/security tools under
 * **`/lexis help` (Shortcut: `/lexis h`)**
   * **What it does:** Displays a quick reference card with all commands, levels, and configuration options.
 
-_(Note: The old individual commands like `/lexis-two-review` are fully supported for backward compatibility but will display a deprecation warning guiding you to use `/lexis` instead.)_
+#### Workflow skills (v1.2)
+
+* **`/lexis doubt`** — Clarify ambiguous requirements (max 3 questions, no code).
+* **`/lexis incremental` (or `inc`)** — Ship the smallest vertical slice; defer the rest.
+* **`/lexis debug` (or `triage`)** — Repro, smallest fix, verify — no drive-by refactors.
+* **`/lexis source` (or `src`)** — Ground in repo code and official docs before inventing APIs.
+* **`/lexis predict`** — Compare proposed vs lazy approach (LOC, deps, maintenance).
+* **`/lexis scenario`** — Happy, edge, and failure scenarios before design.
+
+_(Skill folders remain `skills/lexis-two-*`. Legacy `/lexis-two-*` slash names still work with a deprecation notice.)_
 
 ---
 
@@ -316,10 +325,10 @@ Make it easy to adopt Lexis in any new project.
 - [ ] `AGENTS.template.md` — project-level AGENTS.md template with commented sections (stack, design tokens, glossary, conventions)
 - [x] `docs/portability.md` — hosts, commands, skills, install paths
 - [x] `docs/site.md` — GitHub Pages + `lexis-two.excelso.xyz`
-- [x] `docs/setup.md` — installer guide (phase A1; OpenCode merge in A2)
-- [ ] `docs/modes.md` — when to use lite / full / ultra and how to create custom modes
-- [x] `npx lexis-two install` — setup script (phase A1: rules-only hosts)
-- [ ] README improvements with real `// lexis:` comment examples showing before/after simplifications
+- [x] `docs/setup.md` — installer guide (OpenCode merge + global commands)
+- [x] `docs/modes.md` — lite / full / ultra / off and default resolution
+- [x] `npx @draig/lexis-two install` — setup script (rules, OpenCode, uninstall, hints)
+- [x] README links to `examples/` before/after catalog
 - [x] `examples/nextjs/01-modal-library` — gold-standard before/after (B1)
 - [x] `examples/nextjs/` + `examples/express/` — six before/after cases (B2)
 - [x] `examples/fastapi/` — three before/after cases (B3)
@@ -399,5 +408,9 @@ Focus areas: stack-specific shortcuts for other tech stacks, new examples, addit
 
 MIT — see [LICENSE](./LICENSE).
 
+## Changelog
+
+See [CHANGELOG.md](./CHANGELOG.md).
+
 Original ponytail concept by [DietrichGebert](https://github.com/DietrichGebert/ponytail).
 Lexis ecosystem by [@nitdraig](https://github.com/nitdraig).

package/commands/lexis-two-debug-triage.toml

@@ -0,0 +1,3 @@
+name = "lexis-two-debug-triage"
+description = "Minimal reproduction and smallest bug fix"
+skill = "../skills/lexis-two-debug-triage/SKILL.md"

package/commands/lexis-two-doubt-driven.toml

@@ -0,0 +1,3 @@
+name = "lexis-two-doubt-driven"
+description = "Challenge ambiguous requirements with minimal questions before code"
+skill = "../skills/lexis-two-doubt-driven/SKILL.md"

package/commands/lexis-two-help.toml

@@ -1,3 +1,3 @@
 name = "lexis-two-help"
 description = "Quick reference for lexis-two levels, skills, and commands"
-prompt = "Show the lexis-two quick reference. One shot, change nothing: do not switch mode, write flag files, or persist anything. Levels: /lexis-two lite, /lexis-two (full, default), /lexis-two ultra. Commands: /lexis-two-review, /lexis-two-audit, /lexis-two-debt, /lexis-two-plan, /lexis-two-security, /lexis-two-help."
+prompt = "Show the lexis-two quick reference. One shot, change nothing. Levels: /lexis lite|full|ultra|off. Commands: /lexis plan, review, audit, debt, security, doubt, incremental, debug, source, predict, scenario, help. Legacy lexis-two-* names are deprecated."

package/commands/lexis-two-incremental-impl.toml

@@ -0,0 +1,3 @@
+name = "lexis-two-incremental-impl"
+description = "Ship the smallest vertical slice first"
+skill = "../skills/lexis-two-incremental-impl/SKILL.md"

package/commands/lexis-two-predict.toml

@@ -0,0 +1,3 @@
+name = "lexis-two-predict"
+description = "Compare proposed vs lazy approach — LOC, deps, maintenance"
+skill = "../skills/lexis-two-predict/SKILL.md"

package/commands/lexis-two-scenario.toml

@@ -0,0 +1,3 @@
+name = "lexis-two-scenario"
+description = "Walk happy, edge, and failure scenarios before coding"
+skill = "../skills/lexis-two-scenario/SKILL.md"

package/commands/lexis-two-source-driven.toml

@@ -0,0 +1,3 @@
+name = "lexis-two-source-driven"
+description = "Read official sources and existing code before inventing APIs"
+skill = "../skills/lexis-two-source-driven/SKILL.md"

package/commands/lexis.toml

@@ -1,3 +1,3 @@
 name = "lexis"
 description = "Manage Lexis senior dev mode, intensity levels, and quality/security tools"
-prompt = "Manage Lexis senior dev mode. Handle the requested subcommand: plan (p), review (r), audit (a), debt (d), security (s), help (h), status, or intensity level (lite/full/ultra/off)."
+prompt = "Manage Lexis senior dev mode. Handle the requested subcommand: plan (p), review (r), audit (a), debt (d), security (s), doubt, incremental (inc), debug (triage), source (src), predict, scenario, help (h), status, or intensity level (lite/full/ultra/off)."

package/package.json

@@ -1,58 +1,58 @@
-{
-  "name": "@draig/lexis-two",
-  "version": "1.1.1",
-  "description": "The simple way to obtain the best code. Portable rules, skills, and slash commands for AI agents with lowest tokens usage.",
-  "bin": {
-    "lexis-two": "scripts/install.js"
-  },
-  "main": "./.opencode/plugins/lexis-two.mjs",
-  "exports": {
-    ".": "./.opencode/plugins/lexis-two.mjs",
-    "./server": "./.opencode/plugins/lexis-two.mjs",
-    "./tui": "./.opencode/plugins/lexis-two-tui.mjs"
-  },
-  "oc-plugin": [
-    "server",
-    "tui"
-  ],
-  "files": [
-    "AGENTS.md",
-    ".cursor/rules/lexis-two.mdc",
-    ".windsurf/rules/lexis-two.md",
-    ".clinerules/lexis-two.md",
-    ".kiro/steering/lexis-two.md",
-    ".github/copilot-instructions.md",
-    ".opencode/plugins/lexis-two.mjs",
-    ".opencode/plugins/lexis-two-tui.mjs",
-    ".opencode/commands/",
-    "commands/",
-    "hooks/",
-    "skills/",
-    "templates/",
-    "scripts/",
-    "pi-extension/index.js"
-  ],
-  "keywords": [
-    "pi-package",
-    "pi",
-    "skills",
-    "lexis-two"
-  ],
-  "license": "MIT",
-  "scripts": {
-    "postinstall": "node scripts/opencode-wrapper-postinstall.mjs",
-    "test": "node --test tests/*.test.js && npm test --prefix pi-extension",
-    "benchmark:opencode-go": "node benchmarks/benchmark-opencode-go.js --repeat 3 --write-md",
-    "benchmark:report": "node benchmarks/render-opencode-go-report.js",
-    "site:dev": "npm run dev --prefix site",
-    "site:build": "npm run build --prefix site"
-  },
-  "pi": {
-    "extensions": [
-      "./pi-extension/index.js"
-    ],
-    "skills": [
-      "./skills"
-    ]
-  }
-}
+{
+  "name": "@draig/lexis-two",
+  "version": "1.2.0",
+  "description": "The simple way to obtain the best code. Portable rules, skills, and slash commands for AI agents with lowest tokens usage.",
+  "bin": {
+    "lexis-two": "scripts/install.js"
+  },
+  "main": "./.opencode/plugins/lexis-two.mjs",
+  "exports": {
+    ".": "./.opencode/plugins/lexis-two.mjs",
+    "./server": "./.opencode/plugins/lexis-two.mjs",
+    "./tui": "./.opencode/plugins/lexis-two-tui.mjs"
+  },
+  "oc-plugin": [
+    "server",
+    "tui"
+  ],
+  "files": [
+    "AGENTS.md",
+    ".cursor/rules/lexis-two.mdc",
+    ".windsurf/rules/lexis-two.md",
+    ".clinerules/lexis-two.md",
+    ".kiro/steering/lexis-two.md",
+    ".github/copilot-instructions.md",
+    ".opencode/plugins/lexis-two.mjs",
+    ".opencode/plugins/lexis-two-tui.mjs",
+    ".opencode/commands/",
+    "commands/",
+    "hooks/",
+    "skills/",
+    "templates/",
+    "scripts/",
+    "pi-extension/index.js"
+  ],
+  "keywords": [
+    "pi-package",
+    "pi",
+    "skills",
+    "lexis-two"
+  ],
+  "license": "MIT",
+  "scripts": {
+    "postinstall": "node scripts/opencode-wrapper-postinstall.mjs",
+    "test": "node --test tests/*.test.js && npm test --prefix pi-extension",
+    "benchmark:opencode-go": "node benchmarks/benchmark-opencode-go.js --repeat 3 --write-md",
+    "benchmark:report": "node benchmarks/render-opencode-go-report.js",
+    "site:dev": "npm run dev --prefix site",
+    "site:build": "npm run build --prefix site"
+  },
+  "pi": {
+    "extensions": [
+      "./pi-extension/index.js"
+    ],
+    "skills": [
+      "./skills"
+    ]
+  }
+}

package/pi-extension/index.js

@@ -52,6 +52,47 @@ export function parseLexisCommand(text, defaultMode = DEFAULT_MODE) {
 
 export { writeDefaultMode };
 
+/** @type {Record<string, string>} Maps /lexis subcommand (and short alias) → skill id */
+const LEXIS_SKILL_ROUTES = {
+  plan: "lexis-two-plan",
+  p: "lexis-two-plan",
+  review: "lexis-two-review",
+  r: "lexis-two-review",
+  audit: "lexis-two-audit",
+  a: "lexis-two-audit",
+  debt: "lexis-two-debt",
+  d: "lexis-two-debt",
+  security: "lexis-two-security",
+  s: "lexis-two-security",
+  help: "lexis-two-help",
+  h: "lexis-two-help",
+  doubt: "lexis-two-doubt-driven",
+  incremental: "lexis-two-incremental-impl",
+  inc: "lexis-two-incremental-impl",
+  debug: "lexis-two-debug-triage",
+  triage: "lexis-two-debug-triage",
+  source: "lexis-two-source-driven",
+  src: "lexis-two-source-driven",
+  predict: "lexis-two-predict",
+  scenario: "lexis-two-scenario",
+};
+
+/** @type {Array<{ command: string, lexisSubcommand: string }>} */
+const DEPRECATED_LEXIS_SKILL_COMMANDS = [
+  { command: "lexis-two-review", lexisSubcommand: "review" },
+  { command: "lexis-two-audit", lexisSubcommand: "audit" },
+  { command: "lexis-two-debt", lexisSubcommand: "debt" },
+  { command: "lexis-two-plan", lexisSubcommand: "plan" },
+  { command: "lexis-two-security", lexisSubcommand: "security" },
+  { command: "lexis-two-help", lexisSubcommand: "help" },
+  { command: "lexis-two-doubt-driven", lexisSubcommand: "doubt" },
+  { command: "lexis-two-incremental-impl", lexisSubcommand: "incremental" },
+  { command: "lexis-two-debug-triage", lexisSubcommand: "debug" },
+  { command: "lexis-two-source-driven", lexisSubcommand: "source" },
+  { command: "lexis-two-predict", lexisSubcommand: "predict" },
+  { command: "lexis-two-scenario", lexisSubcommand: "scenario" },
+];
+
 export default function lexisExtension(pi) {
   let currentMode = DEFAULT_MODE;
   let configuredDefaultMode = getDefaultMode();
@@ -95,28 +136,10 @@ export default function lexisExtension(pi) {
         ctx?.ui?.notify?.(`Lexis: current ${currentMode} • default ${configuredDefaultMode}`, "info");
         return;
       }
-      if (subcommand === "plan" || subcommand === "p") {
-        sendAlias("/skill:lexis-two-plan", restArgs, ctx);
-        return;
-      }
-      if (subcommand === "review" || subcommand === "r") {
-        sendAlias("/skill:lexis-two-review", restArgs, ctx);
-        return;
-      }
-      if (subcommand === "audit" || subcommand === "a") {
-        sendAlias("/skill:lexis-two-audit", restArgs, ctx);
-        return;
-      }
-      if (subcommand === "debt" || subcommand === "d") {
-        sendAlias("/skill:lexis-two-debt", restArgs, ctx);
-        return;
-      }
-      if (subcommand === "security" || subcommand === "s") {
-        sendAlias("/skill:lexis-two-security", restArgs, ctx);
-        return;
-      }
-      if (subcommand === "help" || subcommand === "h") {
-        sendAlias("/skill:lexis-two-help", restArgs, ctx);
+
+      const skillId = LEXIS_SKILL_ROUTES[subcommand];
+      if (skillId) {
+        sendAlias(`/skill:${skillId}`, restArgs, ctx);
         return;
       }
 
@@ -148,7 +171,15 @@ export default function lexisExtension(pi) {
     handler: async (args, ctx) => {
       const parsedArgs = String(args || "").trim();
       const [subcommand] = parsedArgs.toLowerCase().split(/\s+/);
-      const validSubcommands = ["status", "plan", "p", "review", "r", "audit", "a", "debt", "d", "security", "s", "help", "h"];
+      const validSubcommands = [
+        "status",
+        ...Object.keys(LEXIS_SKILL_ROUTES),
+        "lite",
+        "full",
+        "ultra",
+        "off",
+        "default",
+      ];
       
       if (validSubcommands.includes(subcommand)) {
         ctx?.ui?.notify?.(`[Deprecated] Use '/lexis ${parsedArgs}' instead.`, "warning");
@@ -164,42 +195,19 @@ export default function lexisExtension(pi) {
     handler: (args, ctx) => sendAlias("/skill:specxis", args, ctx),
   });
 
-  const makeDeprecatedHandler = (subcommand) => {
+  const makeDeprecatedHandler = (lexisSubcommand, skillCommand) => {
     return (args, ctx) => {
-      ctx?.ui?.notify?.(`[Deprecated] Use '/lexis ${subcommand}' instead.`, "warning");
-      sendAlias(`/skill:lexis-two-${subcommand}`, args, ctx);
+      ctx?.ui?.notify?.(`[Deprecated] Use '/lexis ${lexisSubcommand}' instead.`, "warning");
+      sendAlias(`/skill:${skillCommand}`, args, ctx);
     };
   };
 
-  pi.registerCommand("lexis-two-review", {
-    description: "[Deprecated] Use /lexis review instead",
-    handler: makeDeprecatedHandler("review"),
-  });
-
-  pi.registerCommand("lexis-two-audit", {
-    description: "[Deprecated] Use /lexis audit instead",
-    handler: makeDeprecatedHandler("audit"),
-  });
-
-  pi.registerCommand("lexis-two-debt", {
-    description: "[Deprecated] Use /lexis debt instead",
-    handler: makeDeprecatedHandler("debt"),
-  });
-
-  pi.registerCommand("lexis-two-plan", {
-    description: "[Deprecated] Use /lexis plan instead",
-    handler: makeDeprecatedHandler("plan"),
-  });
-
-  pi.registerCommand("lexis-two-security", {
-    description: "[Deprecated] Use /lexis security instead",
-    handler: makeDeprecatedHandler("security"),
-  });
-
-  pi.registerCommand("lexis-two-help", {
-    description: "[Deprecated] Use /lexis help instead",
-    handler: makeDeprecatedHandler("help"),
-  });
+  for (const { command, lexisSubcommand } of DEPRECATED_LEXIS_SKILL_COMMANDS) {
+    pi.registerCommand(command, {
+      description: `[Deprecated] Use /lexis ${lexisSubcommand} instead`,
+      handler: makeDeprecatedHandler(lexisSubcommand, command),
+    });
+  }
 
   pi.on("input", async (event) => {
     if (event?.source === "extension") return;

package/skills/lexis-two-debug-triage/SKILL.md

@@ -0,0 +1,24 @@
+---
+name: lexis-two-debug-triage
+description: Minimal reproduction and smallest fix — no drive-by refactors
+---
+
+Debug the reported issue. **Fix the bug, not the codebase.**
+
+Steps:
+
+1. **Reproduce:** state exact steps or the failing command/output. If not reproducible, say what's missing.
+2. **Hypothesis:** one primary cause (two max if tied).
+3. **Evidence:** file:line or log line — no guessing without reading.
+4. **Fix:** smallest diff that fixes the root cause. No unrelated cleanup.
+5. **Verify:** one command or test that proves the fix.
+
+Forbidden during triage:
+
+- New libraries, new layers, "while we're here" refactors
+- Renaming unrelated symbols
+- Broad error-handler rewrites
+
+If the lazy fix is a one-liner, ship the one-liner. Mark intentional shortcuts with `// lexis:`.
+
+Output: Repro → Cause → Fix (with paths) → Verify. Respond in Spanish.

package/skills/lexis-two-doubt-driven/SKILL.md

@@ -0,0 +1,26 @@
+---
+name: lexis-two-doubt-driven
+description: Challenge ambiguous requirements with minimal questions before any code
+---
+
+The user request is underspecified or scope is unclear. **Do not write code yet.**
+
+Apply YAGNI at the requirement level:
+
+1. Restate what you think they want in one sentence.
+2. Ask at most **3** razor questions that unblock the laziest correct implementation.
+3. For each question, suggest the default you'd pick if they say "whatever is simplest."
+
+Question targets (pick only what applies):
+
+- Does this feature need to exist at all, or does an existing flow cover it?
+- What is the smallest user-visible outcome that counts as done?
+- What can be explicitly out of scope for v1?
+
+Output format:
+
+- **Assumption:** …
+- **Questions:** numbered 1–3
+- **Lazy default if no answer:** one line
+
+No code, no file edits, no new dependencies. Respond in Spanish.

package/skills/lexis-two-incremental-impl/SKILL.md

@@ -0,0 +1,27 @@
+---
+name: lexis-two-incremental-impl
+description: Ship the smallest vertical slice first; defer the rest explicitly
+---
+
+Plan and implement **one** incremental slice — not the full feature.
+
+1. Read the request and list the full feature in one line.
+2. Split into vertical slices ordered by: ships value → reduces risk → deletes unknowns.
+3. Define **slice 1 only**: files touched, behavior delivered, what is explicitly deferred.
+4. Implement slice 1 with the lazy ladder (YAGNI, stdlib, native, existing deps).
+5. End with a **deferred** list (max 5 bullets) — not a roadmap essay.
+
+Rules:
+
+- No scaffolding "for later" in slice 1.
+- No new abstraction until slice 2 proves reuse.
+- If slice 1 is still too big, cut again and say so.
+
+Output:
+
+- **Slice 1 scope:** …
+- **Implementation:** (code or diff as requested)
+- **Deferred:** …
+- **Ship when:** one acceptance check for slice 1
+
+Respond in Spanish.

package/skills/lexis-two-predict/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: lexis-two-predict
+description: Compare proposed approach vs lazy alternative — LOC, deps, maintenance
+---
+
+Compare the user's proposed approach (or the obvious one) against the **lazy alternative**. No implementation unless asked.
+
+For each option (max 2: **proposed** vs **lazy**):
+
+| Dimension | Score |
+| --------- | ----- |
+| New files | count |
+| Est. LOC | range |
+| New dependencies | yes/no — which |
+| Ongoing maintenance | low / med / high — one line why |
+| Failure modes | top 1 |
+
+End with:
+
+- **Pick:** proposed or lazy — one sentence why
+- **Kill condition:** when to revisit (e.g. "third endpoint needs same DTO")
+
+Keep the table compact. No essays. Respond in Spanish.

package/skills/lexis-two-scenario/SKILL.md

@@ -0,0 +1,20 @@
+---
+name: lexis-two-scenario
+description: Walk through concrete scenarios before designing or coding
+---
+
+Before design or code, walk **3 scenarios** for the requested feature:
+
+1. **Happy path** — minimal user action, expected outcome (3–5 lines, concrete data).
+2. **Edge** — boundary input or permission edge (what breaks if ignored).
+3. **Failure** — one realistic failure (network, validation, concurrency) and expected UX/error.
+
+Then:
+
+- **Minimum scope:** what all three scenarios need in common (that's v1).
+- **Defer:** what only edge/failure needs — skip for v1 unless user insists.
+- **Open question:** one line if scenarios conflict.
+
+No code unless the user already asked for implementation. No new dependencies in the plan.
+
+Respond in Spanish.

package/skills/lexis-two-source-driven/SKILL.md

@@ -0,0 +1,22 @@
+---
+name: lexis-two-source-driven
+description: Read official sources and existing code before inventing APIs
+---
+
+Before proposing or implementing, **ground in sources** — not memory.
+
+1. Identify the stack (framework, library versions from `package.json` / lockfile / imports).
+2. Read in-repo truth: types, existing callers, similar patterns (grep/read — don't assume).
+3. Prefer official docs for APIs you will call (link or doc section name).
+4. Propose the smallest approach that matches those sources.
+
+Output:
+
+- **Sources checked:** bullet list (file paths + doc URLs if used)
+- **What exists already:** reuse opportunities
+- **Recommendation:** one paragraph + code only if the user asked for implementation
+- **Avoid:** APIs or patterns not supported by the versions in this repo
+
+If official docs contradict the request, say so and offer the lazy compliant path.
+
+Respond in Spanish.