diff --git a/AGENTS.md b/AGENTS.md
index 90b779d..4641a6e 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -1,17 +1,18 @@
 # AGENTS.md — Guidance for OpenAI Codex
 
 ## Execution Commands
-- **Always** invoke Python scripts via the startup wrapper to activate dependencies:
+- **Always** invoke Python scripts via the startup wrapper, which now installs dependencies globally and then runs your code under the system Python:
   
   ```bash
   ./startup.sh --offset <months> --end <YYYY-MM-DD> --extend-years <N>
+````
 
 ## Project Overview
 
 This repository, **python\_charts**, provides a framework for fetching economic time series from FRED, storing them locally, and producing custom matplotlib charts with configurable parameters. The core components are:
 
-* **`startup.sh`**: Single-entrypoint script to bootstrap a Python venv, install dependencies, fetch data, and run specified commands under the venv interpreter.
-* **`Makefile`**: Convenient high-level targets (`install`, `data`, `plot`, `plot-custom`, `test`, `clean`) that funnel through `startup.sh` for reproducible builds.
+* **`startup.sh`**: Single‐entrypoint script to install dependencies globally, fetch data, and run specified commands under system Python.
+* **`Makefile`**: Convenient high‐level targets (`install`, `data`, `plot`, `plot‐custom`, `test`, `clean`) that funnel through `startup.sh` for reproducible builds.
 * **`scripts/refresh_data.py`**: Downloads UNRATE and DCOILWTICO series into `data/fred.db`.
 * **`scripts/lagged_oil_unrate_chart_styled.py`**: Generates the lagged oil vs. unemployment chart with fixed axes styling and log scale.
 
@@ -21,51 +22,71 @@ Future extensions will add additional custom chart scripts under `scripts/`, all
 
 ### Primary Commands
 
-* **Bootstrap & default plot**: `./startup.sh` or `make plot` — runs the lagged-oil chart with an 18‑month offset.
-* **Custom plot**: `./startup.sh --offset 12 --end 2025-05-31 --extend-years 5` or `make plot-custom ARGS="--offset 12 --end 2025-05-31 --extend-years 5"`.
-* **Data refresh**: `./startup.sh scripts/refresh_data.py` or `make data`.
-* **Run tests**: `./startup.sh pytest -q` or `make test`.
+* **Bootstrap & default plot**:
+
+  ```bash
+  ./startup.sh
+  ```
+
+  Installs/upgrades dependencies globally and runs the default lagged‐oil chart with an 18-month offset.
+
+* **Custom plot**:
+
+  ```bash
+  ./startup.sh --offset 12 --end 2025-05-31 --extend-years 5
+  ```
+
+* **Data refresh**:
+
+  ```bash
+  ./startup.sh scripts/refresh_data.py
+  ```
+
+* **Run tests**:
+
+  ```bash
+  ./startup.sh pytest -q
+  ```
 
 ### Argument Rules
 
-* Arguments beginning with `--` are forwarded to the default chart script.
-* If the first token is a `.py` script or executable, it is invoked under the venv’s Python.
-* Otherwise, commands are executed directly (e.g. linters, test frameworks).
+* If the first token starts with `--`, it’s forwarded to the default chart script.
+* If the first token names a `.py` script, it’s executed via `python <script>`.
+* Otherwise, commands are executed directly (e.g., `pytest`, `flake8`).
 
 ## Coding Standards
 
-* **Python**: Target version 3.12. Use type hints for all function signatures. Follow PEP8 with **Black** formatting.
-* **Bash**: `startup.sh` must use `set -euo pipefail` and always call `$VENV_DIR/bin/python` to avoid system‑Python mismatches.
-* **Makefile**: Use tab‑indented commands. All tasks should funnel to `startup.sh`.
+* **Python**: Target version ≥ 3.8. Use type hints for all function signatures. Follow PEP8 with Black formatting.
+* **Bash**: `startup.sh` uses `set -euo pipefail`.
+* **Makefile**: Use tab‐indented commands. All targets should funnel through `startup.sh`.
 
 ## Testing & Validation
 
-* **Automated tests**: No current tests; add pytest modules under `tests/` following the pattern `test_*.py`.
+* **Automated tests**: Add pytest modules under `tests/` following the pattern `test_*.py`.
 * **Data mocking**: For new chart scripts, mock FRED downloads via monkeypatching `pandas_datareader.data.DataReader`.
 
 ## Extending with New Charts
 
-1. **Create a new script** under `scripts/`, for example `scripts/custom_chart.py`.
-2. **Implement** a `main()` accepting `--start`, `--end`, `--offset` or custom args using `argparse`.
-3. **Leverage** `startup.sh` invocation: add usage example in your script’s docstring.
-4. **Update** the `Makefile` if you need a dedicated target, otherwise users can call:
+1. **Create** `scripts/custom_chart.py`.
+2. **Implement** a `main()` with `argparse` for `--start`, `--end`, `--offset`, etc.
+3. **Invoke** via:
 
    ```bash
    ./startup.sh python scripts/custom_chart.py --your-args
    ```
+4. **Optionally** add a Makefile target if desired.
 
 ## CI / Codex Integration
 
-* In Codex cloud or CLI tasks, set the **Setup Script** to `./startup.sh install data` to ensure dependencies and data are ready.
-* **Task Commands** should call `make plot` or `make test`, which delegates to `startup.sh`.
-* **AGENTS.md** is automatically loaded by Codex to understand structure and commands.
+* **Setup Script** →
 
-## PR & Commit Guidelines
+  ```bash
+  chmod +x startup.sh && ./startup.sh
+  ```
+* **Task Command** →
 
-* Prefix branches/PRs with `feature/` or `fix/`.
-* Commit messages: `<type>(<scope>): <description>` (e.g. `feat(charts): add custom inflation plot`).
-* Include screenshots or saved chart outputs under `outputs/` for visual verification.
-
----
+  ```bash
+  ./startup.sh --offset 18
+  ```
 
 *End of AGENTS.md — tailored for python\_charts repo*
diff --git a/Makefile b/Makefile
index df06693..f07bb19 100644
--- a/Makefile
+++ b/Makefile
@@ -1,29 +1,26 @@
-# Makefile for python_charts
-# Defines standard targets for Codex/CI to use startup.sh
+.PHONY: install data plot plot-custom test clean
 
-.PHONY: install data plot test clean
-
-# Install dependencies into venv and prepare data
+# Install dependencies & fetch data
 install:
-	./startup.sh pip install
+	./startup.sh
 
 # Download FRED data (if needed)
 data:
 	./startup.sh scripts/refresh_data.py
 
-# Generate the default chart (18-month lag)
+# Default 18-month lag plot
 plot:
 	./startup.sh
 
-# Generate a custom chart	
-# Usage: make plot ARGS="--offset 12 --end 2025-05-31 --extend-years 5"
+# Custom plot
+# Usage: make plot-custom ARGS="--offset 12 --end 2025-05-31 --extend-years 5"
 plot-custom:
 	./startup.sh $(ARGS)
 
-# Run test suite
+# Run tests
 test:
 	./startup.sh pytest -q
 
-# Remove virtual environment and outputs
+# Clean up any outputs (no venv to delete)
 clean:
-	rm -rf venv data outputs
+	rm -rf outputs
diff --git a/README.md b/README.md
index 77b2f03..7c0a33b 100644
--- a/README.md
+++ b/README.md
@@ -1,12 +1,10 @@
-# Lagged Oil–Unemployment Chart
+# Python Charts
 
-A self-contained toolkit that:
+This initial chart is a self-contained toolkit that:
 
-* Fetches U.S. unemployment rate (UNRATE) and WTI crude oil price (DCOILWTICO) from FRED
-* Stores the series in a local SQLite database (`data/fred.db`)
-* Produces a dual-axis plot of unemployment vs. lagged oil price, with customizable lag, date range, and styling
-
-![Example chart](scripts/lagged_oil_unrate_chart_styled.py)
+* Fetches U.S. unemployment rate (UNRATE) and WTI crude oil price (DCOILWTICO) from FRED  
+* Stores the series in a local SQLite database (`data/fred.db`)  
+* Produces a dual-axis plot of unemployment vs. lagged oil price, with customizable lag, date range, and styling  
 
 ---
 
@@ -18,17 +16,19 @@ git clone https://github.com/your-org/python_charts.git
 cd python_charts
 
 # One-command: install deps, download data, and plot
-./startup.sh python scripts/lagged_oil_unrate_chart_styled.py \
-    --offset 12 --end 2025-05-31 --extend-years 5
-```
+./startup.sh --offset 12 --end 2025-05-31 --extend-years 5
+````
 
 This will:
 
-1. Create (or reuse) a Python 3 virtual environment (`venv/`)
-2. Upgrade `pip` & `setuptools` (providing `distutils` support)
-3. Install required packages (offline wheelhouse first, PyPI fallback)
-4. Download UNRATE & DCOILWTICO into `data/fred.db` if missing
-5. Generate the chart with a 12-month oil lead, extending the x-axis by 5 years
+1. Upgrade `pip` & `setuptools` globally
+2. Install required packages from `requirements.txt` via
+
+   ```bash
+   pip install --upgrade -r requirements.txt
+   ```
+3. Download UNRATE & DCOILWTICO into `data/fred.db` if missing
+4. Generate the chart with a 12-month oil lead, extending the x-axis by 5 years
 
 ---
 
@@ -37,46 +37,28 @@ This will:
 ```
 python_charts/
 ├── data/
-│   └── fred.db                 # Local FRED snapshot (SQLite)
+│   └── fred.db                            # Local FRED snapshot (SQLite)
 ├── scripts/
-│   ├── refresh_data.py         # Download and store FRED series
+│   ├── refresh_data.py                    # Download and store FRED series
 │   ├── lagged_oil_unrate_chart_styled.py  # Plotting script
-│   └── pyproject.toml          # Packaging metadata
-├── startup.sh                  # Bootstraps venv, deps, data, and runs commands
-├── requirements.txt            # Python dependencies
-└── README.md                   # This file
+│   └── pyproject.toml                     # Packaging metadata
+├── Makefile                               # High-level targets (install, plot, etc.)
+├── startup.sh                             # Installs deps globally, fetches data, runs commands
+├── requirements.txt                       # Python dependencies
+└── README.md                              # This file
 ```
 
 ---
 
-## 🛠️ Installation & Setup
-
-1. **Ensure** you have Python 3.8+ installed on your system.
-2. **Clone** and navigate into the project:
-
-   ```bash
-   git clone https://github.com/your-org/python_charts.git
-   cd python_charts
-   ```
-3. **Make** the startup script executable:
-
-   ```bash
-   chmod +x startup.sh
-   ```
-
-> On first run, `startup.sh` will build the environment, install dependencies, and pull data.
-
----
-
 ## ⚙️ Usage
 
 ### Refresh FRED data
 
 ```bash
-./startup.sh python scripts/refresh_data.py
+./startup.sh scripts/refresh_data.py
 ```
 
-Downloads the full UNRATE and WTI series (from 1948 to today) into `data/fred.db`.
+Downloads UNRATE and WTI series (from 1948 to today) into `data/fred.db`.
 
 ### Generate the chart
 
@@ -84,17 +66,16 @@ Downloads the full UNRATE and WTI series (from 1948 to today) into `data/fred.db
 # Default 18-month lag
 ./startup.sh
 
-# Custom lag, range, and extension
-./startup.sh python scripts/lagged_oil_unrate_chart_styled.py \
-    --offset 6 --start 1980-01-01 --end 2025-05-31 --extend-years 2
+# Custom lag, date range, and extension
+./startup.sh --offset 6 --start 1980-01-01 --end 2025-05-31 --extend-years 2
 ```
 
-| Option           | Description                                               |
-| ---------------- | --------------------------------------------------------- |
-| `--offset`       | Lag in months (positive = oil leads; negative = oil lags) |
-| `--start`        | Start date (YYYY-MM-DD)                                   |
-| `--end`          | End date (YYYY-MM-DD)                                     |
-| `--extend-years` | Years to extend the x-axis beyond the end date            |
+| Option           | Description                                     |
+| ---------------- | ----------------------------------------------- |
+| `--offset`       | Lag in months (pos = oil leads; neg = oil lags) |
+| `--start`        | Start date (YYYY-MM-DD)                         |
+| `--end`          | End date (YYYY-MM-DD)                           |
+| `--extend-years` | Years to extend the x-axis beyond the end date  |
 
 ---
 
@@ -105,7 +86,7 @@ Downloads the full UNRATE and WTI series (from 1948 to today) into `data/fred.db
 
   * Left y-axis: unemployment rate (%) from 3 to 15%, ticks every 2%.
   * Right y-axis: log scale oil price (USD), doubling ticks from \$3 upward.
-  * X-axis: years, major tick every 5 years, minor tick every year.
+  * X-axis: major tick every 5 years, minor tick every year.
 
 ---
 
@@ -121,18 +102,17 @@ lxml>=4.6.0
 python-dateutil>=2.8.1
 ```
 
-`startup.sh` also upgrades `pip` & `setuptools` to ensure `distutils` is available.
-
 ---
 
 ## 🛡️ Troubleshooting
 
-* **ModuleNotFoundError: distutils**: resolved automatically by upgrading `setuptools` in `startup.sh`.
-* **Database not found**: run `./startup.sh python scripts/refresh_data.py` to populate `data/fred.db`.
-* **Virtual-env issues**: delete `venv/` and re-run `startup.sh`.
+* **ModuleNotFoundError: distutils**: resolved by `pip install --upgrade setuptools`.
+* **Database not found**: run `./startup.sh scripts/refresh_data.py`.
+* **Permission errors**: ensure write access to the project directory.
 
 ---
 
 ## 📄 License
 
-This project is released under the MIT License. See [LICENSE](LICENSE) for details.
+Released under the MIT License. See [LICENSE](LICENSE) for details.
+
diff --git a/repo-context.txt b/repo-context.txt
index afecdf7..c7c9c3c 100644
--- a/repo-context.txt
+++ b/repo-context.txt
@@ -25,6 +25,12 @@ Generated on: 2025-05-30
 ```markdown
 # AGENTS.md — Guidance for OpenAI Codex
 
+## Execution Commands
+- **Always** invoke Python scripts via the startup wrapper to activate dependencies:
+  
+  ```bash
+  ./startup.sh --offset <months> --end <YYYY-MM-DD> --extend-years <N>
+
 ## Project Overview
 
 This repository, **python\_charts**, provides a framework for fetching economic time series from FRED, storing them locally, and producing custom matplotlib charts with configurable parameters. The core components are:
diff --git a/startup.sh b/startup.sh
index 91981e2..0927018 100755
--- a/startup.sh
+++ b/startup.sh
@@ -1,69 +1,48 @@
 #!/usr/bin/env bash
 # -------------------------------------------------------------------
-# startup.sh  — prepare Python env, ensure deps & data, then run.
+# startup.sh  — install deps globally, fetch data, then run command
 # Usage:
-#   ./startup.sh [options]      # passes options to default plot script
-#   ./startup.sh [script or cmd] # runs specified script/command
+#   ./startup.sh [options]       # passes options to default plot script
+#   ./startup.sh <script> [args] # runs specified script with system Python
 # Examples:
 #   ./startup.sh --offset 12 --end 2025-05-31 --extend-years 5
 #   ./startup.sh pytest -q
 # -------------------------------------------------------------------
 set -euo pipefail
 
-VENV_DIR="venv"
 REQ_FILE="requirements.txt"
-WHEELHOUSE="wheelhouse"
 DATA_DB="data/fred.db"
-PYTHON="$VENV_DIR/bin/python"
 DEFAULT_SCRIPT="scripts/lagged_oil_unrate_chart_styled.py"
 
-# 1) Create or reuse venv
-if [[ ! -d "$VENV_DIR" ]]; then
-  echo "🔧 Creating virtual-env…"
-  python3 -m venv "$VENV_DIR"
-fi
-
-# 2) Activate venv
-# shellcheck disable=SC1090
-source "$VENV_DIR/bin/activate"
-
-# 3) Upgrade pip & setuptools
+# 1) Ensure pip & setuptools are up to date
 echo "🛠 Upgrading pip & setuptools…"
-$PYTHON -m pip install --upgrade pip setuptools
+pip install --upgrade pip setuptools
 
-# 4) Install dependencies
-PIP_OPTS=(install --upgrade --no-input)
-if [[ -d "$WHEELHOUSE" ]]; then
-  echo "📦 Installing from wheelhouse…"
-  $PYTHON -m pip "${PIP_OPTS[@]}" --no-index --find-links="$WHEELHOUSE" -r "$REQ_FILE" \
-    || { echo "⚠️ Wheelhouse failed—falling back…"; \
-         $PYTHON -m pip "${PIP_OPTS[@]}" -r "$REQ_FILE"; }
-else
-  echo "🌐 Installing from PyPI…"
-  $PYTHON -m pip "${PIP_OPTS[@]}" -r "$REQ_FILE"
-fi
+# 2) Install project dependencies globally
+echo "📦 Installing dependencies from $REQ_FILE…"
+pip install --upgrade -r "$REQ_FILE"
 
-echo "✅ Virtual-env ready ($($PYTHON -V))"
-
-# 5) Ensure data exists
+# 3) Ensure data exists
 if [[ ! -f "$DATA_DB" ]]; then
   echo "🔄 Data not found—downloading FRED series…"
-  $PYTHON scripts/refresh_data.py
+  python scripts/refresh_data.py
 fi
 
-# 6) Build command
+# 4) Dispatch to the right command
 if [[ $# -eq 0 ]]; then
-  CMD=("$PYTHON" "$DEFAULT_SCRIPT" --offset 18)
+  # no args → run default plot
+  CMD=(python "$DEFAULT_SCRIPT" --offset 18)
 elif [[ "$1" == --* ]]; then
-  CMD=("$PYTHON" "$DEFAULT_SCRIPT" "$@")
+  # args start with -- → forward to default plot
+  CMD=(python "$DEFAULT_SCRIPT" "$@")
 else
+  # first token is script or other command
   if [[ -f "$1" || "$1" == *.py ]]; then
-    CMD=("$PYTHON" "$@")
+    CMD=(python "$@")
   else
     CMD=("$@")
   fi
 fi
 
-# 7) Execute
 echo "🚀 Executing: ${CMD[*]}"
 exec "${CMD[@]}"