atla

Unified Atlassian CLI for Jira and Confluence.

atla is a Rust single-binary CLI for day-to-day Atlassian Cloud work. It provides Jira and Confluence commands with profile-based authentication, human-friendly tables, and machine-friendly output formats.

Why atla?

• Work with Jira issues, boards, and sprints without leaving the terminal.
• Read and update Confluence spaces, pages, blogs, labels, comments, and attachments from scripts or interactive workflows.
• Automate reporting and CI with stable table, json, csv, and keys output, mutation policies (--dry-run, --read-only), and context budgets.

Install

Verified installer script (macOS/Linux):

base=https://github.com/NeoHsu/atla/releases/latest/download
curl --proto '=https' --tlsv1.2 -LsSfO "$base/atla-installer.sh"
curl --proto '=https' --tlsv1.2 -LsSfO "$base/atla-installer.sh.sha256"
shasum -a 256 -c atla-installer.sh.sha256
sh atla-installer.sh

Verified Windows PowerShell installer:

$base = "https://github.com/NeoHsu/atla/releases/latest/download"
Invoke-WebRequest "$base/atla-installer.ps1" -OutFile atla-installer.ps1
Invoke-WebRequest "$base/atla-installer.ps1.sha256" -OutFile atla-installer.ps1.sha256
$expected = ((Get-Content -Raw atla-installer.ps1.sha256).Trim() -split '\s+')[0]
$actual = (Get-FileHash -Algorithm SHA256 atla-installer.ps1).Hash.ToLowerInvariant()
if ($actual -ne $expected) { throw "installer checksum verification failed" }
& .\atla-installer.ps1

Direct release downloads:

Platform	Asset
Apple Silicon macOS	atla-aarch64-apple-darwin.tar.xz
Intel macOS	atla-x86_64-apple-darwin.tar.xz
ARM64 Linux	atla-aarch64-unknown-linux-gnu.tar.xz
x64 Linux	atla-x86_64-unknown-linux-gnu.tar.xz
x64 Windows	atla-x86_64-pc-windows-msvc.zip

Each archive contains the prebuilt atla executable plus README and license files. Checksums, build-provenance attestations, and a CycloneDX 1.5 binary SBOM (atla.cdx.xml plus atla.cdx.xml.sha256) are published with the release assets; release notes come from the changelog.

mise:

mise use -g github:NeoHsu/atla

In mise.toml:

[tools]
"github:NeoHsu/atla" = "latest"

From source:

cargo install --git https://github.com/NeoHsu/atla atla

Common workflows

printf '%s\n' "$ATLASSIAN_TOKEN" | atla auth login --instance https://example.atlassian.net \
  --email you@example.com --token-stdin
atla jira search "assignee = currentUser() order by updated desc" --limit 10
atla jira issue transition PROJ-123 --to Done --dry-run
atla plan jira issue create --project PROJ --type Task --summary 'Agent task' --out issue-plan.json
atla apply issue-plan.json --yes --output json
atla confluence page view 67890 --format markdown
atla confluence search "type=page AND space=DEV" --output json

Documentation

File	Description
Getting Started	Installation, first-time setup, shell completions, quick demo
Authentication	Auth commands, multi-profile management, token storage, env vars
Configuration	Config keys, aliases, config.toml schema, environment overrides
Jira	Full Jira command reference: projects, issues, boards, sprints
Confluence	Full Confluence command reference: spaces, pages, blogs, attachments
Output Formats	Global flags, output formats, scripting/CI patterns
JSON Contracts	Versioned schemas, plans, receipts, compatibility policy
Saved Plans	Generate and safely apply expiring mutation plans
Operation Policy	Read-only enforcement, allow/deny rules, context budgets
Config Migration	Schema v2 migration, backups, and rollback
Compatibility	Supported platforms and v1 stability policy
Release Procedure	Artifacts, SBOM, provenance, and workflow hardening
Contributing	Development checks, CLI contracts, and PR process
Code Generation	Progenitor-based API client generation, spec filtering, update workflow
Agent Reference	Structured command reference for AI agents and automation

atla-cli Skill

Install the bundled AI agent skill to enable atla-aware assistance — JQL/CQL patterns, all command flags, scripting idioms, and safety rules.

From the published repository:

npx skills add NeoHsu/atla --skill atla-cli

From a local checkout of this repo, install the internal skills/atla-cli package with:

npx skills add . --skill atla-cli

For non-interactive setup across all supported agents, add --agent '*' -y:

npx skills add . --skill atla-cli --agent '*' -y

Use --copy if you want the installed skill to be a standalone copy instead of a symlink back to the repo checkout.

Feature Matrix

Product	Resource	Commands
Core	Auth	login, logout, status, switch
Core	Config	set, get, list
Jira	Projects	list, view, issue-types
Jira	Search	JQL search with table, JSON, CSV, and key output
Jira	Issues	list, create, view, update, edit, delete
Jira	Issue fields	fields — list create-meta fields with required flag, type, and allowed values
Jira	Assignment	assign --to me, account ID, or user query
Jira	Transitions	List or apply transitions, with interactive selection when possible
Jira	Comments	comment add, comment list, comment update, comment delete
Jira	Attachments	attachment list, attachment upload, attachment download, attachment delete
Jira	Links	link add, link list, link remove, link github-links, link github-commits
Jira	Worklogs	worklog add, worklog list
Jira	Boards	board list with project, type, and name filters; board view
Jira	Sprints	sprint list, sprint active, sprint view, sprint create, sprint start, sprint close, sprint add, sprint remove, sprint issues
Confluence	Spaces	list, view, create, update, delete
Confluence	Pages	list, view, create, update, move, delete, children, copy
Confluence	Page content	Storage, wiki, Atlas Doc Format input; storage, ADF, markdown view output
Confluence	Page labels	label list, label add, label remove
Confluence	Page comments	comment add, comment list, comment delete
Confluence	Blogs	list, view, create, update, delete
Confluence	Blog labels	label list, label add, label remove
Confluence	Blog comments	comment add, comment list, comment delete
Confluence	Search	CQL search through scoped v1 REST endpoint
Confluence	Attachments	list, view, upload, download, delete
Output	Formats	table, json, csv, keys
Safety	Mutation policy	Global --dry-run previews and --read-only enforcement
Safety	Context budgets	--max-pages, --max-items, --max-bytes, --timeout

Configuration

atla stores profile configuration in ~/.config/atla/config.toml by default. API tokens are stored through the OS keyring and are not written to config files.

Initial auth:

printf '%s\n' "$ATLASSIAN_TOKEN" | atla auth login --instance https://example.atlassian.net \
  --email you@example.com --token-stdin
atla auth status
atla config set default-project PROJ
atla config set alias.mine "jira search 'assignee = currentUser() order by updated desc'"
atla config list --output json

atla stores API tokens in the OS keyring by default. Atlassian tokens have a configurable expiration of 1–365 days, so rotate them before expiry and re-run atla auth login when needed. Unscoped tokens use the site URL. For a scoped token, pass the tenant cloud ID; atla then routes Jira and Confluence through their respective api.atlassian.com/ex/{product}/{cloudId} gateways while retaining the site URL for web links:

printf '%s\n' "$ATLASSIAN_TOKEN" | atla auth login --instance https://example.atlassian.net \
  --cloud-id 11111111-2222-3333-4444-555555555555 --email you@example.com --token-stdin

In headless or container environments where keyring access is unavailable, use an explicit file-backed credential store or provide a token through the environment:

printf '%s\n' "$ATLASSIAN_TOKEN" | atla auth login --storage file \
  --instance https://example.atlassian.net --email you@example.com --token-stdin
ATLA_TOKEN="$ATLASSIAN_TOKEN" atla jira project list

File-backed credentials are stored separately from the main config in ~/.config/atla/credentials.toml by default. Override paths with ATLA_CONFIG and ATLA_CREDENTIALS for isolated runs.

Aliases expand before command parsing, so the alias above can be used as:

atla mine --limit 25

Shell completions:

atla completion bash > ~/.local/share/bash-completion/completions/atla
atla completion zsh > ~/.zfunc/_atla
atla completion fish > ~/.config/fish/completions/atla.fish

Usage

Jira examples:

atla jira project list
atla jira project list --query platform --limit 25 --output json
atla jira project view PROJ
atla jira search "project = PROJ order by created desc"
atla jira issue list --project PROJ --status "In Progress"
atla jira issue create --project PROJ --type Task --summary "Fix login"
atla jira issue update PROJ-123 --summary "Updated summary"
atla jira issue edit PROJ-123 --labels add:urgent,remove:low
atla jira issue view PROJ-123 --web
atla jira issue assign PROJ-123 --to me
atla jira issue transition PROJ-123 --to Done
atla jira issue comment add PROJ-123 "Ready for review"
atla jira issue comment add PROJ-123 "Please check the screenshot" --attachment ./screenshot.png
atla jira issue delete PROJ-123 --yes
atla jira issue fields --project PROJ --type Bug --required-only
atla jira board list --project PROJ
atla jira sprint active --board 84

Confluence examples:

atla confluence space list
atla confluence space create "Development" --key DEV --description "Team docs"
atla confluence page list --space DEV
atla confluence page view 67890 --format markdown
atla confluence page view 67890 --format markdown --preserve-table-options
atla confluence page create --space DEV --title "Meeting Notes" --body-file notes.html
atla confluence page create --space DEV --title "Inventory" --body-file inventory.md --representation markdown --numbered-table-rows
atla confluence page create --space DEV --title "Runbook" --body-file runbook.md --representation markdown --mention "Neo Hsu=abc-account-id"
atla confluence page update 67890 --title "Updated Notes"
atla confluence page move 67890 --parent 12345
atla confluence page label add 67890 runbook urgent
atla confluence page comment add 67890 "Looks good"
atla confluence page comment add 67890 "Please review the report" --representation markdown --attachment ./report.pdf
atla confluence blog create --space DEV --title "Release Notes" --body-file release.html
atla confluence search "type=page AND space=DEV"
atla confluence attachment upload 67890 ./diagram.png --comment "Updated diagram"
atla confluence attachment download 13579 --save-to ./diagram.png

Confluence v2 remains the primary generated client. Confluence search, attachment upload, and page label mutation use scoped Confluence v1 REST endpoints where v2 does not expose the required operation.

Pagination

List and search commands treat --limit N as the maximum number of records to return for that invocation. If more records exist, table output prints a ready-to-copy next command:

More results available.
Next page:
  atla jira project list --limit 25 --page-token <TOKEN>

JSON output includes the same information under pagination.nextPageToken and pagination.nextCommand. The token is opaque and validated against the command/query that created it. Use --all instead of --limit when you intentionally want to fetch every matching record. For agent runs, bound context and network work globally:

atla --read-only --max-pages 5 --max-items 200 --max-bytes 1000000 --timeout 30 \
  --output json jira search 'project = PROJ ORDER BY updated DESC'

When --max-pages or --max-items stops an --all request, atla emits a resume token instead of pretending the result is complete. --read-only rejects local and remote mutations before credentials or network access.

Development

mise trust
mise install
cargo check
cargo test

Refresh Atlassian specs:

scripts/update-specs.sh
cargo check --workspace

Override config paths for isolated test runs:

ATLA_CONFIG=/tmp/atla-config.toml cargo run -p atla -- config list

Release

Releases are automated via cargo-dist. Pushing a version tag builds Linux, macOS, and Windows binaries and publishes them to GitHub Releases.