jirascrap

A terminal UI for browsing your Jira tickets. Built with Bubble Tea.

Tickets are fetched from the Jira API and cached locally in SQLite. You can tag tickets, manage per-ticket todo lists, filter, and search -- all from the terminal.

Requirements

• Go 1.22+
• A Jira Cloud instance with API access

Installation

git clone https://github.com/codedogapp/jirascrap.git
cd jirascrap
go build -o jirascrap .

Configuration

Set these environment variables before running:

Variable	Required	Description
JIRA_BASE_URL	Yes	Your Jira instance URL (e.g. https://yourorg.atlassian.net)
JIRA_EMAIL	Yes	Email associated with your Atlassian account
JIRA_API_TOKEN	Yes	API token from Atlassian API tokens
JIRA_DB_PATH	No	Path to SQLite database file (default: ./data/jira.db)
JIRASCRAP_LOG_DIR	No	Directory for session log files (default: ~/.local/state/jirascrap/logs/)
JIRASCRAP_COPILOT_WORKSPACE	No	Directory where Copilot CLI launches (default: current working directory)
JIRASCRAP_COPILOT_MODEL	No	AI model for initial planning (default: claude-haiku-4.5)

Usage

jirascrap

Key bindings

Key	Action
enter	Select ticket
esc	Go back / close popup
t	Tag current ticket
n	Open todo list
s	Change ticket status
r	Refresh tickets from Jira
o	Open ticket in browser
c	Send ticket to Copilot CLI (tmux)
?	Toggle full help
q	Quit
ctrl+c	Force quit
/	Filter tickets

In the tag popup:

Key	Action
tab	Autocomplete tag
up/down	Navigate suggestions
enter	Save tags
esc	Cancel

In the todo popup:

Key	Action
a	Add new todo
space	Toggle done
x	Delete todo
esc	Close

How it works

On startup, jirascrap loads any cached tickets from the local SQLite database and displays them immediately. In the background, it fetches fresh data from the Jira API, updates the cache, and refreshes the UI. This means the app is usable instantly, even on slow connections.

Tags and todos are stored locally and are preserved even if a ticket is removed from your Jira query results.

Press r at any time to manually sync with Jira.

Logging

Session logs are written to ~/.local/state/jirascrap/logs/ (configurable via JIRASCRAP_LOG_DIR). Each app session creates a timestamped log file. The 10 most recent log files are kept; older ones are pruned automatically.

Copilot CLI Integration

Press c on any ticket to launch GitHub Copilot CLI in a tmux session with full ticket context.

How it works:

• A single tmux session named copilot is created (or reused)
• Each ticket gets its own pane, identified by ticket ID
• Copilot starts in plan mode with a cheap model and the ticket context as the initial prompt
• The TUI stays running — switch to the copilot session with tmux attach -t copilot

Requirements: tmux and copilot must be in your PATH.

Configuration:

Variable	Default	Description
JIRASCRAP_COPILOT_WORKSPACE	Current directory	Working directory for Copilot
JIRASCRAP_COPILOT_MODEL	claude-haiku-4.5	Model for initial planning

Development

Run the app:

go run .

Run tests:

go test ./...

Run the e2e demo (requires vhs, ttyd, and ffmpeg):

bash e2e/run.sh