JIRA Ticket Prioritizer

implement

forge

Arguments

• $ARGUMENTS[0]

  — Comma-separated ticket keys (e.g.

  EC-100,EC-101,EC-102

  ) OR a JQL query (detected by presence of

  =

  ,

  AND

  ,

  OR

  )

• $ARGUMENTS[1]

  — (optional) Additional context for prioritization (e.g. "focus on backend", "frontend first")

System Requirements

• jira

  CLI installed and configured (https://github.com/ankitpokhrel/jira-cli)
• Environment variable

  JIRA_API_TOKEN

  set with a valid Jira API token. Important: When checking this variable, verify at least 2 times before concluding it is not set. Never expose the value — use existence checks only.

Execution

Step 1 — Parse Input

• Check if

  $ARGUMENTS[0]

  contains

  =

  ,

  AND

  , or

  OR

  (case-insensitive) to detect JQL
• If JQL detected: run

  jira issue list --jql "$ARGUMENTS[0]" --plain --columns key --no-headers

  to resolve to ticket keys
• If comma-separated: split on

  ,

  and trim whitespace
• Validate each key matches

  [A-Z]+-\d+

• If fewer than 2 valid tickets, report the single ticket and exit

Step 2 — Fetch All Tickets

• Create temp directory:

  mkdir -p .jira-ticket-prioritizer-tmp/tickets

• For each ticket key, fetch via jira CLI:

  jira issue view {KEY} --raw > .jira-ticket-prioritizer-tmp/tickets/{KEY}.json

• Parse each ticket using the jira-ticket-viewer parse script:

  node ../jira-ticket-viewer/scripts/parse-ticket.js < .jira-ticket-prioritizer-tmp/tickets/{KEY}.json > .jira-ticket-prioritizer-tmp/tickets/{KEY}-parsed.json

• Collect all parsed outputs into a single JSON array and write to

  .jira-ticket-prioritizer-tmp/all-tickets.json

Step 3 — Build Dependency Graph

• Run:

  node ./scripts/build-dependency-graph.js < .jira-ticket-prioritizer-tmp/all-tickets.json > .jira-ticket-prioritizer-tmp/graph.json

• Review graph output for cycles or warnings
• See references/dependency-analysis.md for relationship mapping rules

Step 4 — Semantic Dependency Analysis

• For tickets flagged with

  "relates to"

  links or no explicit links in the graph output, analyze their

  description

  ,

  summary

  , and

  commentSummary

  for implicit dependencies
• Look for:
  • References to other ticket keys in the input set
  • Shared component or module mentions
  • Functional ordering hints ("builds on", "requires", "after", "depends on")
• Add soft edges with confidence levels (high/medium/low) to the graph
• Re-run topological sort if new edges were added:
  • Update

    all-tickets.json

    with additional

    softEdges

    and re-run

    build-dependency-graph.js

Step 5 — Priority Scoring

• For tickets at the same dependency layer, score using weights from references/priority-weights.md
• Skip tickets with status "Done" — list them as excluded in the report
• Sort within each layer by descending score
• Apply

  $ARGUMENTS[1]

  context as a tiebreaker or boost (e.g. "focus on backend" boosts tickets with backend labels/components)

Step 6 — Generate Output

• Write

  .jira-ticket-prioritizer-tmp/detailed-report.json

  — full details including scores, justifications, dependency graph, excluded tickets, and warnings (see references/output-format.md for schema)
• Write

  .jira-ticket-prioritizer-tmp/output.json

  — ticket keys grouped by dependency layer (array of arrays), sorted by priority score within each layer
• Present only

  output.json

  (the simple sorted array) to the user. The detailed report is saved for reference but not displayed.

Reference Files