Issue Summary

ACLI exposes a fixed set of Jira subcommands that cover specific REST operations. When a user needs to call a Jira REST endpoint not yet covered by a dedicated subcommand, there is no escape hatch, they must fall back to curl or direct REST calls, losing ACLI's managed authentication, site context, and output formatting.

This forces teams to maintain parallel tooling (ACLI + raw REST), preventing full standardization on ACLI as the single CLI interface for Jira automation.

Customer Use Case

A customer evaluating ACLI as their standard Jira CLI cannot fully adopt it because several REST capabilities they depend on are not yet exposed. Examples from customer use case:

1. Upload attachments to work items → acli jira workitem attachment supports list and delete but not upload, blocking CI/CD pipelines from attaching build artifacts

1. Use issue picker-style search endpoints → for autocomplete/fuzzy-match workflows beyond standard JQL

1. Retrieve paginated issue changelogs → for audit and compliance reporting requiring full change history

1. Look up project components → for automated issue routing and triage

1. Resolve allowed users for assignee fields → for scripts that validate and set assignees programmatically

1. Discover field metadata and custom field IDs → for dynamic scripting that adapts to different Jira configurations

Rather than waiting for each to be implemented individually, a generic pass-through command using ACLI's existing authenticated context would unblock adoption immediately.

Prior Art

This pattern is established in other CLI tools:

• GitHub CLI — GitHub CLI api calls any GitHub REST/GraphQL endpoint using the authenticated context

• Buildkite CLI — bk api provides the same escape hatch

Expected Results

1. Add acli jira api subcommand → specify HTTP method, endpoint path, query params, and request body using the existing authenticated session

1. Support common REST patterns → GET, POST, PUT, DELETE; JSON bodies; pagination helpers; file upload for multipart endpoints

1. Preserve ACLI output formatting → structured output (table, JSON, plain) so it integrates into scripted pipelines

1. If the subcommand cannot be prioritized short-term, consider accelerating coverage of the specific gaps listed above (especially attachment upload and field metadata discovery) as individual subcommands