Azure pipelines tool
Skill lbussell/AzurePipelinesTool/skills/azure-pipelines-tool
Interact with Azure Pipelines from the terminal. Use when you need to list pipelines, inspect definitions, validate YAML, queue runs, monitor status, or download logs. Run from within a Git repo that has Azure DevOps pipelines.From its SKILL.md
npx -y skills add lbussell/AzurePipelinesTool --skill azure-pipelines-toolAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
3.7 KB, as published. Nobody here has run it
azure-pipelines-cli
A CLI tool for interacting with Azure Pipelines.
Requirements
- .NET 10+ SDK
- Authenticated via
azd auth login(uses Azure Developer CLI credential) - Run from within a Git repo whose remote points to Azure DevOps
Installation
Use dnx to run without global installation (like npx for Node):
dnx azp -y -- <command>
Important:
- Always use
-yto skip the interactive confirmation prompt (which breaks LLM tool use). - Always use
--to separate dnx options from tool arguments.
Workflows
Commands chain together in two natural flows. Each step's output feeds into the next.
Pipeline Development: list → info → check → run → wait
# 1. Discover pipelines in the current repo (shows name, ID, YAML path)
dnx azp -y -- list
# 2. Inspect a pipeline's variables and parameters
dnx azp -y -- info path/to/pipeline.yml
# 3. Validate YAML expansion with template parameters (dry run, no queue)
dnx azp -y -- check path/to/pipeline.yml --parameters env=staging
# 4. Queue the run with multiple parameters, variables, and stage skips
dnx azp -y -- run path/to/pipeline.yml --parameters env=staging,imageTag=latest --variables tag=v1,debug=true -s Deploy,Cleanup
# 5. Wait for completion using the build ID from step 4
dnx azp -y -- wait 12345 -f
check and run require a clean working tree synced with upstream — commit and push first.
Monitoring: status → logs
# 1. View run status as a tree (stages → jobs → tasks)
dnx azp -y -- status 12345 -d 3
# 2. Download logs for a specific task (logId shown in status -d 3 output)
dnx azp -y -- logs 12345 42
# Cancel a running build
dnx azp -y -- cancel 12345
status and logs also accept full Azure DevOps build URLs instead of numeric IDs.
Key Flags
| Flag | Purpose | Commands |
|---|---|---|
--parameters k=v[,k=v,...] | Template parameter overrides (comma-separated) | check, run |
--variables k=v[,k=v,...] | Pipeline variable overrides (comma-separated; must be settable at queue time) | run |
-s/--skip stage[,stage,...] | Stage names to skip (comma-separated) | run |
-d 1|2|3 | Tree depth: 1=stages, 2=+jobs (default), 3=+tasks | status |
-f | Exit with non-zero code on failure/cancellation | wait |
Command Reference
| Command | Purpose |
|---|---|
list | List all pipelines associated with the current repository |
info <path> | Show pipeline details: variables, parameters, metadata |
check <path> | Preview expanded YAML (dry run with template parameters) |
run <path> | Queue a pipeline run with parameters, variables, and stage skips |
status <id> | Show run status as a tree of stages, jobs, and tasks |
cancel <id> | Cancel a running pipeline build |
wait <id> | Poll until a run completes, with optional failure exit code |
logs <id> <logId> | Download logs for a specific task from a run |
llmstxt | Print comprehensive tool documentation |
install-skill | Install the agent skill to a local or user directory |
<path> is a relative path to the pipeline YAML file.
<id> is a numeric build ID or a full Azure DevOps build results URL.
When to Use This Skill
- Listing and inspecting Azure Pipelines definitions in a repo
- Validating pipeline YAML before queuing a run
- Queuing pipeline runs with custom parameters and variables
- Monitoring build progress and downloading task logs
- Canceling in-progress pipeline runs