| name | task-api-helper |
| description | Helper skill for teams using the shared Task API REST service and its task_cli.py wrapper. Use when the user needs to interact with project tasks: list tasks, filter them by status, fetch task details, add comments, and work within the current shared Task API workflow. The skill knows the API contract, CLI command set, and the improvement process for proposing upstream enhancements when teams discover repetitive or awkward workflows. |
| license | MIT |
| allowed-tools | Bash,Python |
task-api-helper
Overview
This skill helps you work with the Task API — a shared REST service for project task management — and its companion CLI wrapper task_cli.py. Use this skill when you need to:
- Query or update project tasks via the REST API
- Run the
task_cli.py wrapper commands
- Understand the current shared Task API workflow
- Navigate the improvement process to propose a new command upstream
Environment Setup
cp .env.example .env
TASK_API_URL="https://tasks.internal.example.com"
TASK_API_TOKEN="<your-token>"
The installed skill reads configuration in this order:
--api-url / --token
.env in the installed skill folder
- shell environment variables
Using .env in the installed skill folder is the preferred demo setup because
the agent may execute in a different shell than the one where environment
variables were exported.
Supported CLI Commands
list-tasks
python task_cli.py list-tasks [--status <status>] [--api-url <url>]
Lists all tasks. Use --status to filter by any status value the API supports, for example waiting-for-response or resolved. Omit --status to return all tasks.
get-task
python task_cli.py get-task TASK_ID [--api-url <url>]
Fetches a single task by ID including its comment thread.
add-comment
python task_cli.py add-comment TASK_ID "Your comment text" [--api-url <url>]
Appends a comment to a single task. The comment text is a positional argument.
Workflow evolution
This skill is intentionally centrally maintained. If you discover that a common
workflow feels repetitive, awkward, or missing a useful higher-level command,
do not jump straight to an upstream issue.
First:
- try a temporary local proof of concept in the current repo
- verify whether it actually helps
- capture the exact command or behavior you tested
- remove the local patch
Only then open an enhancement issue in the catalog repository using the
Task API Enhancement template. See
references/IMPROVEMENT-PROCESS.md for the full flow.
Quick Diagnostics
python task_cli.py --help
python task_cli.py list-tasks --status waiting-for-response
Improvement Requests
This skill is centrally maintained. Proposed enhancements should be backed by a
short local proof of concept in the consumer repo before an upstream issue is
opened. Reference the issue number in any PR that implements the change. See
references/IMPROVEMENT-PROCESS.md for the full lifecycle.