gh-agent-task
Skill4 starsChanged 6 months agoMITView on GitHub
GitHub CLI (`gh agent-task`) operations for creating, listing, and viewing preview agent tasks. You MUST load this skill when working with the `gh agent-task` command.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- github/gh-awtheir own version, first seen May 27, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- github/gh-aw★ 5.6ksame name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository124 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .github/AGENTS.mdAGENTS.md
- .github/workflows/AGENTS.mdAGENTS.md
- .vscode/AGENTS.mdAGENTS.md
- .github/copilot-instructions.mdCopilot instructions
- agent-log-analysisSkill
- agent-mdSkill
- agent-md-writerSkill
- agent-skill-md-writerSkill
- agentskillsSkill
- agents-md-writerSkill
- ai-prompt-writerSkill
- ansibleSkill
- apache-airflow-apiSkill
- apache-airflow-dagsSkill
- astro-cliSkill
- astronomer-docsSkill
- awesome-copilotSkill
- brainstorm-agent-runsSkill
- brainstorm-github-prSkill
- brainstormSkill
- catSkill
- chrome-devtoolsSkill
- claude-docsSkill
- codeqlSkill
- code-reviewSkill
- code-tourSkill
- coding-standard-writerSkill
- context-aware-opsSkill
- copilot-cliSkill
- copilot-docsSkill
- critical-thinkingSkill
- datadog-agentSkill
- datadog-apiSkill
- datadog-mcpSkill
- datadog-monitorsSkill
- datadog-pulumiSkill
- devcontainerSkill
- dictationSkill
- direnvSkill
- dockerfileSkill
- dockerSkill
- docs-reviewSkill
- docs-writerSkill
- dot-claudeSkill
- dotfilesSkill
- dot-githubSkill
- editorconfigSkill
- fact-writerSkill
- gdpr-compliantSkill
- gh-apiSkill
- gh-aw-compileSkill
- gh-aw-firewall-debugSkill
- gh-aw-firewallSkill
- gh-aw-newSkill
- gh-awSkill
- gh-aw-troubleshootingSkill
- gh-codespaceSkill
- gh-issueSkill
- gh-modelsSkill
- The first 60 shown.
The file
Show its rules and text
We split this skill into the rules it gives an agent
13 rules. About 0 are common, 4 are distinctive, and 9 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
9
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 16
When you need to create, view, or list GitHub Copilot preview agent tasks directly from the CLI.
- Distinctive: line 18
When investigating the status or logs of an asynchronous agent task session.
- Very specific: line 23
If your environment uses an older version of gh (< v2.80.0) that doesn't support the agent-task plugin.
- Very specific: line 40
Keep descriptions clear when creating tasks.
- Very specific: line 41
Always use non-interactive mode.
- Very specific: line 47
Create a task from a file (recommended for complex tasks): gh agent-task create -F task-desc.md
- Very specific: line 50
Create a task using stdin: echo "description" | gh agent-task create -F -
- Very specific: line 53
Create a task with a custom agent on a specific base branch: gh agent-task create "fix errors" --base main --custom-agent my-agent
- Very specific: line 56
Create a task and follow logs: gh agent-task create "build me a new app" --follow
- Very specific: line 83
If a task creation fails, verify repository permissions and ensure the GitHub CLI is authenticated (gh auth status).
- Very specific: line 84
If using a custom agent, ensure the agent file exists in .github/agents/.
- Distinctive: line 98
gh: Use for general GitHub CLI operations.
- Distinctive: line 99
gh-pr: Use for pull request specific operations.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Skill: gh-agent-task
<!-- markdownlint-disable MD013 MD023 MD031 MD032 -->Provides usage patterns and expert guidance for the GitHub CLI gh agent-task extension, which manages preview agent tasks on repositories and pull requests.
The agent-task command set is only available in v2.80.0 or later of the GitHub CLI.
This command set is a public preview and is subject to change.
When to Use
- When you need to create, view, or list GitHub Copilot preview agent tasks directly from the CLI.
- To automate the assignment of an agent to a specific pull request or issue.
- When investigating the status or logs of an asynchronous agent task session.
When Not to Use
- For standard GitHub CLI operations (like listing PRs or Issues) that don't involve the agent task extension.
- If your environment uses an older version of
gh(< v2.80.0) that doesn't support theagent-taskplugin. - For triggering generic GitHub Actions (use
gh workflow runinstead).
Common Pitfalls
- Interactive Prompts: Running
gh agent-task createwithout-For providing an inline description, causing the CLI to hang waiting for user input. - Parsing Raw Text: Relying on grep to parse the task status instead of using the built-in
--jsonand--jqflags. - Assuming Completion: Firing off an agent task but failing to poll or
--followits status, leading to race conditions in automated scripts.
Core Process
- Determine if the goal is to create a new task, list existing tasks, view a specific task, or view tasks associated with a PR.
- Select the correct subcommand (
create,list, orview). - Apply filtering (
--jq), formatting (--json,--template), or input options (-F) as required.
Core Principles
- Keep descriptions clear when creating tasks; prefer using a file (
-F) for multi-line instructions instead of inline strings. - Always use non-interactive mode. For JSON extraction, rely on the built-in
--jqand--jsonflags rather than external piping when possible.
Commands / Usage Patterns
Create an Agent Task
Create a task from a file (recommended for complex tasks):
gh agent-task create -F task-desc.md
Create a task using stdin:
echo "description" | gh agent-task create -F -
Create a task with a custom agent on a specific base branch:
gh agent-task create "fix errors" --base main --custom-agent my-agent
Create a task and follow logs:
gh agent-task create "build me a new app" --follow
List Agent Tasks
List tasks and format output with JSON/JQ:
gh agent-task list --limit 10 --json id,state,pullRequestNumber --jq '.[] | "\(.id): \(.state)"'
View Agent Tasks
View an agent task by session ID:
gh agent-task view e2fa49d2-f164-4a56-ab99-498090b8fcdf
View an agent task by PR number:
gh agent-task view 12345
View an agent task by PR reference:
gh agent-task view OWNER/REPO#12345
View task and output specific JSON fields:
gh agent-task view 12345 --json id,state,pullRequestNumber
Show agent session logs:
gh agent-task view 12345 --log
Diagnostics and Troubleshooting
- If a task creation fails, verify repository permissions and ensure the GitHub CLI is authenticated (
gh auth status). - If using a custom agent, ensure the agent file exists in
.github/agents/.
Limitations
- The
gh agent-taskcommands are currently in preview and subject to change.
References
Related Skills
- gh: Use for general GitHub CLI operations.
- gh-pr: Use for pull request specific operations.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

