stg <- dev

README.md

@@ -52,7 +52,7 @@ For direct OpenAI runs you can also use your local Codex browser login instead o
 npx skill-optimizer init cli       # or: init sdk, init mcp, init prompt
 ```
 
-The wizard asks for your repo path, models to benchmark, and where your `SKILL.md` lives. It creates a `skill-optimizer/` directory:
+The wizard asks for your repo path, models to benchmark, and where your `SKILL.md` lives. It creates a `.skill-optimizer/` directory:
 - `skill-optimizer.json` — the main config (commit this)
 - `.skill-optimizer/cli-commands.json` — CLI surface manifest (template to edit, or auto-extracted)
 - `.skill-optimizer/tools.json` — MCP surface manifest (template to edit)
@@ -68,13 +68,13 @@ npx skill-optimizer import-commands --from my-cli --scrape
 **Step 3 — Run a benchmark:**
 
 ```bash
-npx skill-optimizer run --config ./skill-optimizer/skill-optimizer.json
+npx skill-optimizer run --config ./.skill-optimizer/skill-optimizer.json
 ```
 
 **Step 4 — Run the optimizer** (iteratively improves your `SKILL.md`):
 
 ```bash
-npx skill-optimizer optimize --config ./skill-optimizer/skill-optimizer.json
+npx skill-optimizer optimize --config ./.skill-optimizer/skill-optimizer.json
 ```
 
 The optimizer never modifies your original `SKILL.md` — it works from versioned local copies in `.skill-optimizer/` and prints a progress table at the end showing per-model improvement.
@@ -113,7 +113,7 @@ npx skill-optimizer init --answers answers.json
 }
 ```
 
-**Key config fields** in `skill-optimizer/skill-optimizer.json`:
+**Key config fields** in `.skill-optimizer/skill-optimizer.json`:
 
 | Field | What it does | Set it to |
 |-------|-------------|-----------|

SKILL/SKILL.md

@@ -39,7 +39,7 @@ Before doing anything, figure out where you are:
 | Run optimizer | `npx skill-optimizer optimize --config <config-path>` |
 | Compare two runs | `npx skill-optimizer compare --baseline a.json --current b.json` |
 
-`<config-path>` is the path to your `skill-optimizer.json` — typically `./skill-optimizer/skill-optimizer.json` after running `init`, or wherever you placed it.
+`<config-path>` is the path to your `skill-optimizer.json` — typically `./.skill-optimizer/skill-optimizer.json` after running `init`, or wherever you placed it.
 
 ## What Do You Need?
 

SKILL/references/setup.md

@@ -78,7 +78,7 @@ npx skill-optimizer init --answers answers.json
 {
   "surface": "cli",
   "repoPath": "/absolute/path/to/your-repo",
-  "models": ["openrouter/anthropic/claude-sonnet-4.6", "openrouter/openai/gpt-4o"],
+  "models": ["openrouter/anthropic/claude-sonnet-4.6", "openrouter/openai/gpt-4o-mini"],
   "maxTasks": 20,
   "maxIterations": 5,
   "entryFile": "src/cli.ts"
@@ -130,7 +130,7 @@ npx skill-optimizer doctor --fix --config <config-path>
 
 After successful setup:
 
-- **`skill-optimizer.json`** — main config file (commit this); when created by `init`, the default location is `./skill-optimizer/skill-optimizer.json`
+- **`skill-optimizer.json`** — main config file (commit this); when created by `init`, the default location is `./.skill-optimizer/skill-optimizer.json`
 - **`.skill-optimizer/`** — working directory for task artifacts, surface manifests, and versioned skill copies (gitignored)
 
 Your project is ready for benchmarking. Read `references/benchmark.md` for next steps.