Jira Issue Data Extractor

A secure, CLI-based toolset for pulling issue data from Jira. The repo contains two independent entry points:

Script	Purpose
main.py	Extract issues for any JQL query, compute time metrics, and run local AI theme analysis via Ollama. Writes a CSV and a Markdown report.
jira_metrics_report.py	Print a quick terminal metrics summary scoped to a fixed set of assignees — active issue counts, weekly averages, and wait-time breakdown. No files are written.

What It Extracts (main.py)

• Issue title, assignee, account name, Rancher team, reporter
• Created / updated / resolved date-time
• Current status, type, priority, resolution
• Description and comments
• Issue key (for traceability)
• Status change history (used for time-in-status metrics)

Quick Start

1. Create a virtual environment and install dependencies

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

2. Configure credentials

cp .env.example .env

Jira Server / Data Center:

• JIRA_BASE_URL=https://jira.suse.com
• JIRA_API_TOKEN=<Personal Access Token> — create at Profile → Personal Access Tokens in Jira.

---

main.py — Theme Analysis & Time Metrics

Runs a JQL query, fetches the full issue set (including status changelog for time-in-status metrics), writes a CSV, and produces a Markdown report with AI-generated themes via a local Ollama instance.

Run

python3 main.py 'project IN ("SURE", "NVSHAS") AND created >= -7d AND issuetype in ("Bug", "Escalation")'

Output

Timestamped files written to output/:

• jira_issues_<timestamp>.csv — raw issue data
• jira_issue_themes_<timestamp>.md — Markdown report containing:
  • Status breakdown — closed vs. open counts
  • Time metrics (always shown):
    • Time to initial response — creation to first comment by the assignee
    • Time actively worked — open time excluding Waiting for Reporter periods
    • Time waiting for reporter — total time spent in Waiting for Reporter status
  • Resolution statistics (shown when JQL targets resolved issues):
    • Average and median time to resolution
  • Issues by customer and team
  • AI-generated themes, cross-cutting observations, and team-specific action items

If Ollama is unreachable or returns invalid JSON, extraction still completes and the tool prints a "Theme analysis skipped" message.

CLI options

python3 main.py \
  --timeout 1200 \
  --ollama-retries 3 \
  --page-title 'Weekly Escalation Report' \
  --base-filename escalations_weekly \
  'project IN ("SURE", "NVSHAS") AND created >= -7d AND issuetype in ("Bug", "Escalation")'

Flag	Default	Description
--timeout	900	Seconds per Ollama request
--ollama-retries	2	Retries for transient Ollama failures (timeouts, temporary connectivity, 5xx/429)
--page-title	"Jira Issue Theme Analysis"	Heading in the Markdown report
--base-filename	"jira_issue_themes"	Base name for output files

Environment variables

Variable	Default	Description
JIRA_EXTRACTOR_OUTPUT_DIR	output	Directory for output files
OLLAMA_ENABLED	true	Set to false to skip AI analysis
OLLAMA_MODEL	llama3.1:8b	Ollama model to use
OLLAMA_HOST	http://localhost:11434	Ollama endpoint (OLLAMA_BASE_URL is also accepted as a fallback)

---

jira_metrics_report.py — Assignee Metrics Dashboard

Prints a terminal summary for a fixed set of assignees: active issue counts, issues waiting for reporter, and weekly new/resolved averages. No files are written. Requires JIRA_ASSIGNEES to be set.

Run

JIRA_ASSIGNEES="jane.doe,john.smith" python3 jira_metrics_report.py

Or set JIRA_ASSIGNEES in your .env file and run:

python3 jira_metrics_report.py

CLI options

Flag	Default	Description
--weeks	12	Rolling window for new/resolved weekly averages
--projects	SURE,NVSHAS	Comma-separated Jira project keys

Example:

python3 jira_metrics_report.py --weeks 8 --projects SURE,NVSHAS,OBS

Required environment variable

Variable	Description
JIRA_ASSIGNEES	Comma-separated list of Jira usernames to scope the report to

---

jira_phrase_extractor.py — Phrase/Window Issue Extractor

Standalone phrase-based extractor with no AI theme analysis. It builds JQL from --phrase and --window, fetches matching issues, and writes CSV output.

Run

# Last year (default: 1y)
python3 jira_phrase_extractor.py --phrase "MetalLB"

# Last 9 months
python3 jira_phrase_extractor.py --phrase "MetalLB" --window 9m

# Last 2 years
python3 jira_phrase_extractor.py --phrase "MetalLB" --window 2y

# Scope to specific projects
python3 jira_phrase_extractor.py --phrase "MetalLB" --window 1y --projects SURE,NVSHAS

# Only bug issues
python3 jira_phrase_extractor.py --phrase "MetalLB" --window 2y --type Bug

# Multiple issue types
python3 jira_phrase_extractor.py --phrase "MetalLB" --window 2y --type Bug --type Escalation

Generated JQL looks like:

text ~ "\"MetalLB\"" AND created >= "2025-06-04"

The extractor intentionally uses an absolute cutoff date for compatibility with Jira instances that reject relative year/month period syntax like -2y.

Switch the date field for window filtering:

python3 jira_phrase_extractor.py --phrase "MetalLB" --window 9m --date-field updated

CLI options

Flag	Default	Description
--phrase	(required)	Phrase to search for in issue text.
--window	1y	Time window (1y, 9m, 2y, 12w, 30d, etc.).
--date-field	created	Date field used for the window (created or updated).
--projects	(unset)	Optional comma-separated project keys (SURE,NVSHAS).
--type	(unset)	Optional issue type filter; repeat for multiple values (--type Bug --type Escalation).
--base-filename	jira_phrase_issues	Base filename for generated CSV output.

---

Notes

• Jira Server / DC uses POST /rest/api/2/search with startAt/total pagination.
• main.py fetches the status changelog per issue (GET /rest/api/2/issue/{key}/changelog) to compute time-in-status metrics. This adds one extra API call per issue.
• The "Account Name" field is hardcoded to customfield_23901.
• The "Rancher Team" field is hardcoded to customfield_23900.
• AI analysis runs only against your local Ollama endpoint. No issue data is sent to any hosted LLM service.

Finding the Rancher Team custom field id

curl -sS \
  -H "Authorization: Bearer ${JIRA_API_TOKEN}" \
  -H "Accept: application/json" \
  "${JIRA_BASE_URL}/rest/api/2/field" | jq -r '.[] | select(.name == "Rancher Team") | .id'