jon23d/issue-tracker-read

Issue Tracker: Read

Reads a ticket from the configured issue tracker and returns structured context.

Setup

Read agent-config.json from the repository root to determine the provider:

cat agent-config.json

Use the issue_tracker.provider field to select the correct commands below.

---

GitHub

Required

• gh CLI installed and authenticated

Read a ticket

gh issue view <ticket-id> --json number,title,body,labels,assignees,milestone,state \
  --repo <issue_tracker.github.repo_url>

Extract fields

From the JSON response, extract:

• title — issue title
• body — full description (contains requirements, acceptance criteria)
• labels[].name — label names (use for scope hints)
• state — open/closed

---

Gitea

Required

• GITEA_TOKEN environment variable set
• GITEA_URL environment variable set (e.g. https://gitea.example.com)
• Owner and repo extracted from issue_tracker.gitea.repo_url

Read a ticket

curl -s \
  -H "Authorization: token $GITEA_TOKEN" \
  -H "Content-Type: application/json" \
  "$GITEA_URL/api/v1/repos/<owner>/<repo>/issues/<ticket-id>"

Extract fields

From the JSON response, extract:

• title
• body
• labels[].name
• state

---

Jira

Required

• JIRA_TOKEN environment variable set (API token)
• JIRA_EMAIL environment variable set
• JIRA_URL environment variable set (e.g. https://yourorg.atlassian.net)

Read a ticket

curl -s \
  -u "$JIRA_EMAIL:$JIRA_TOKEN" \
  -H "Content-Type: application/json" \
  "$JIRA_URL/rest/api/3/issue/<ticket-id>"

Extract fields

Jira responses are nested. Extract:

• fields.summary → title
• fields.description.content → description (Atlassian Document Format — extract plain text from text nodes)
• fields.labels → label names
• fields.status.name → current status
• fields.acceptanceCriteria → may be a custom field, check fields for keys containing "acceptance"

---

Output

After reading the ticket, return a structured context object — do not pass the raw API response downstream:

{
  "id": "47",
  "title": "One-line summary",
  "description": "Full requirement text",
  "labels": ["backend", "auth"],
  "raw_acceptance_criteria": "Text of acceptance criteria if present in ticket body"
}

If acceptance criteria are not explicitly stated in the ticket, note that in raw_acceptance_criteria as null — the orchestrator will infer them during extraction.