AI Jira Summarizer

This project generates summaries of Jira issues and posts them back into the Jira issue description

Required environment variables

It is recommended that you provide the variables via a .env file since they will contain sensitive information. There is already a .gitignore file that will ignore the .env file.

The following variables are required:

• ALLOWED_PROJECTS: A comma-separated list of Jira project keys that the bot is allowed to summarize (e.g., ALLOWED_PROJECTS=OCTO,OCTOET) (for the bot)
• CONFLUENCE_TOKEN: A Confluence PAT token that will allow updating the Confluence pages with the summaries (for rollup_status.py)
• CONFLUENCE_URL: The URL for the Confluence instance (e.g., https://...)
• GENAI_API: The API endpoint for the IBM AI model (e.g., https://...)
• GENAI_KEY: Your API key for the IBM AI model
• JIRA_TOKEN: A JIRA PAT token that will allow retrieving issues from Jira as well as posting the AI summaries
• JIRA_URL: The URL for the Jira instance (e.g., https://...)
• JWT_SECRET_KEY: The secret key for the JWT token (for the API)

Commands

Summarize a single Jira issue: summarize_issue.py

$ pipenv run ./summarize_issue.py --help
Loading .env environment variables...
usage: summarize_issue.py [-h] [-d MAX_DEPTH] [--log-level
            {DEBUG,INFO,WARNING,ERROR,CRITICAL}] [-n] [-r] jira_issue_key

Summarize a JIRA issue

positional arguments:
  jira_issue_key        JIRA issue key

options:
  -h, --help            show this help message and exit
  -d MAX_DEPTH, --max-depth MAX_DEPTH
                        Maximum depth to recursively examine issues while summarizing
  --log-level {DEBUG,INFO,WARNING,ERROR,CRITICAL}
                        Set the logging level
  -n, --no-update       Do not update the Jira issues with the summaries
  -r, --regenerate      Force regeneration of summaries

Summarizer bot: bot.py

$ pipenv run ./bot.py --help
Loading .env environment variables...
usage: bot.py [-h] [-d MAX_DEPTH] [--log-level
            {DEBUG,INFO,WARNING,ERROR,CRITICAL}] [-m MODIFIED_SINCE] [-n] [-s SECONDS]

Summarizer bot

options:
  -h, --help            show this help message and exit
  -d MAX_DEPTH, --max-depth MAX_DEPTH
                        Maximum depth to recursively examine issues while summarizing
  --log-level {DEBUG,INFO,WARNING,ERROR,CRITICAL}
                        Set the logging level
  -m MODIFIED_SINCE, --modified-since MODIFIED_SINCE
                        Summarize issues that have been modified since this date/time
  -n, --no-update       Do not update the Jira issues with the summaries
  -s SECONDS, --seconds SECONDS
                        Seconds to wait between iterations

Methods to run the summarizer bot

Note: All these methods assume that the required environment variables are configured via a .env file, as described above.

• Run locally via pipenv:
  • Install pipenv: pip install pipenv
  • Install the dependencies: pipenv install
  • Run the bot: pipenv run ./bot.py <bot-args>
• Run locally via Docker:
  • Build the Docker image: docker build -t jira-summarizer .
  • Run the Docker container: docker run --env-file .env jira-summarizer <bot-args>
• Run in Kubernetes:
  • Create a Namespace for the project: kubectl create namespace jira-summarizer
  • Create a Kubernetes secret with the environment variables: kubectl -n jira-summarizer create secret generic jira-summarizer-secret --from-env-file=.env
  • Apply the Kubernetes manifest: kubectl -n jira-summarizer apply -f jira-summarizer.yaml You may need to adjust the image specification in the manifest to point to your Docker image repository

Summarizer API: summarize_api.py

Generate JWT tokens for users to access the summarizer API:

$ pipenv run ./summarize_api.py
Loading .env environment variables...
Usage: ./summarize_api.py <userid>

Run locally for development/testing:

$ pipenv run flask --app summarize_api.py run
Loading .env environment variables...
 * Tip: There are .env or .flaskenv files present. Do "pip install
   python-dotenv" to use them.
 * Serving Flask app 'summarize_api.py'
 * Debug mode: off
WARNING: This is a development server. Do not use it in a production deployment.
Use a production WSGI server instead.
 * Running on http://127.0.0.1:5000
Press CTRL+C to quit

The API can then be tested via curl:

$ curl -s -H "Authorization: Bearer YOUR_GENERATED_API_TOKEN" \
http://127.0.0.1:5000/api/v1/summarize-issue?key=PROJ-1234 | jq . -r
{
  "key": "PROJ-1234",
  "stats": {
    "fetchrelated_time": 1.436011292,
    "getissue_time": 0.284554751,
    "llm_time": 5.364057366,
    "request_time": 7.088276041
  },
  "summary": "This is the issue summary.",
  "user": "zzz"
}

To run in OpenShift:

• Create the Namespace and Secret as described above
• Apply the Kubernetes manifest: kubectl -n jira-summarizer apply -f summarize-api.yaml