Add Makefile, some simple readme info. Remove python2 mentioning

There are no changes which mean python2 is actually unsupportd, but we don't want to guarantee this going forwards so remove it from the README

.gitignore

@@ -0,0 +1 @@
+targets

Makefile

@@ -0,0 +1,8 @@
+.PHONY: format
+format:
+	poetry run black .
+
+.PHONY: dev-start
+dev-start:
+	@mkdir -p ./targets
+	poetry run python discoverecs.py --directory $$PWD/targets

README.md

@@ -1,62 +1,92 @@
 # prometheus-ecs-sd
+
 ECS Service Discovery for Prometheus
 
 ## Info
+
 This tool provides Prometheus service discovery for Docker containers running on AWS ECS. You can easily instrument your app using a Prometheus
 client and enable discovery adding an ENV variable at the Service Task Definition. Your container will then be added
-to the list of Prometheus targets to be scraped. Requires python2 or python3 and boto3. Works with Prometheus 2.x. It supports bridge, host
-and awsvpc (EC2 and Fargate) network modes.
+to the list of Prometheus targets to be scraped.
+
+Bridge, host and awsvpc (EC2 and Fargate) network modes are supported.
+
+Requires
+
+-   `python3`
+-   The `boto3` library
+-   Prometheus `2.x`.
+
+## Developing
+
+Local development requires the [poetry](https://python-poetry.org/) tool, check the documentation for installation instructions.
+
+To start the service run
+
+```shell
+AWS_PROFILE=<your_aws_profile> make dev-start
+```
+
+any AWS configuration supported by boto3 is supported, e.g. individual access/secret keys.
+
+To format code run
+
+```shell
+make format
+```
 
 ## Setup
+
 `discoverecs.py` should run alongside the Prometheus server. It generates targets using JSON file service discovery. It can
 be started by running:
 
-```
+```shell
 python discoverecs.py --directory /opt/prometheus-ecs
 ```
 
-Where `/opt/prometheus-ecs` is defined in your Prometheus config as a `file_sd_config` job:
+note that the directory must already exist.
+
+The output directory is then `/opt/prometheus-ecs` defined in your Prometheus config as a `file_sd_config` job:
 
 ```yaml
 - job_name: 'ecs-1m'
   scrape_interval: 1m
   file_sd_configs:
-    - files:
-        - /opt/prometheus-ecs/1m-tasks.json
+      - files:
+            - /opt/prometheus-ecs/1m-tasks.json
   relabel_configs:
-    - source_labels: [metrics_path]
-      action: replace
-      target_label: __metrics_path__
-      regex: (.+)
+      - source_labels: [metrics_path]
+        action: replace
+        target_label: __metrics_path__
+        regex: (.+)
 ```
 
 You can also specify a discovery interval with `--interval` (in seconds). Default is 60s. We also provide caching to minimize hitting query
 rate limits with the AWS ECS API. `discoverecs.py` runs in a loop until interrupted and will output target information to stdout.
 
 To make your application discoverable by Prometheus, you need to set the following environment variable in your task definition:
 
-```
-{"name": "PROMETHEUS", "value": "true"}
+```json
+{ "name": "PROMETHEUS", "value": "true" }
 ```
 
 Metric path and scrape interval is supported via `PROMETHEUS_ENDPOINT`:
 
-```
+```text
 "interval:/metric_path,..."
 ```
 
 Examples:
 
-```
+```text
 "5m:/mymetrics,30s:/mymetrics2"
 "/mymetrics"
 "30s:/mymetrics1,/mymetrics2"
 ```
 
 Under ECS task definition (`task.json`):
 
-```
-{"name": "PROMETHEUS_ENDPOINT", "value": "5m:/mymetrics,30s:/mymetrics2"}
+```json
+{ "name": "PROMETHEUS_ENDPOINT", "value": "5m:/mymetrics,30s:/mymetrics2" }
 ```
 
 Available scrape intervals: `15s`, `30s`, `1m`, `5m`.
@@ -66,50 +96,51 @@ The default metric path is `/metrics`. The default scrape interval is `1m`.
 The following Prometheus configuration should be used to support all available intervals:
 
 ```yaml
-  - job_name: 'ecs-15s'
-    scrape_interval: 15s
-    file_sd_configs:
+- job_name: 'ecs-15s'
+  scrape_interval: 15s
+  file_sd_configs:
       - files:
-          - /opt/prometheus-ecs/15s-tasks.json
-    relabel_configs:
+            - /opt/prometheus-ecs/15s-tasks.json
+  relabel_configs:
       - source_labels: [metrics_path]
         action: replace
         target_label: __metrics_path__
         regex: (.+)
 
-  - job_name: 'ecs-30s'
-    scrape_interval: 30s
-    file_sd_configs:
+- job_name: 'ecs-30s'
+  scrape_interval: 30s
+  file_sd_configs:
       - files:
-          - /opt/prometheus-ecs/30s-tasks.json
-    relabel_configs:
+            - /opt/prometheus-ecs/30s-tasks.json
+  relabel_configs:
       - source_labels: [metrics_path]
         action: replace
         target_label: __metrics_path__
         regex: (.+)
 
-  - job_name: 'ecs-1m'
-    scrape_interval: 1m
-    file_sd_configs:
+- job_name: 'ecs-1m'
+  scrape_interval: 1m
+  file_sd_configs:
       - files:
-          - /opt/prometheus-ecs/1m-tasks.json
-    relabel_configs:
+            - /opt/prometheus-ecs/1m-tasks.json
+  relabel_configs:
       - source_labels: [metrics_path]
         action: replace
         target_label: __metrics_path__
         regex: (.+)
-        
-  - job_name: 'ecs-5m'
-    scrape_interval: 5m
-    file_sd_configs:
+
+- job_name: 'ecs-5m'
+  scrape_interval: 5m
+  file_sd_configs:
       - files:
-          - /opt/prometheus-ecs/5m-tasks.json
-    relabel_configs:
+            - /opt/prometheus-ecs/5m-tasks.json
+  relabel_configs:
       - source_labels: [metrics_path]
         action: replace
         target_label: __metrics_path__
         regex: (.+)
 ```
+
 ## EC2 IAM Policy
 
 The following IAM Policy should be added when running `discoverecs.py` in EC2:
@@ -135,6 +166,7 @@ resource "aws_iam_role_policy_attachment" "prometheus-server-role-ec2-read-only"
 ```
 
 ## Special cases
+
 For skipping labels, set `PROMETHEUS_NOLABELS` to `true`.
 This is useful when you use "blackbox" exporters or Pushgateway in a task
 and metrics are exposed at a service level. This way, no EC2/ECS labels