agentsclimarketplace

Issue to agent

Skill itscloud0/issue-to-agent/.github/skills/issue-to-agent

Turn GitHub issues into ready-to-run task packs for coding agents.

Install
npx -y skills add itscloud0/issue-to-agent --skill issue-to-agent

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 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.

What its author says it does

Copied from the file, not written here

Generate a ready-to-run implementation task pack from a GitHub issue and repository context for Copilot agents.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

3.5 KB, 790 tokens by cl100k_base, as published. Nobody here has run it

issue-to-agent

Task Trigger

Use this skill when a user asks GitHub Copilot, VS Code agent mode, Copilot CLI, or Copilot cloud agent to:

  • prepare a GitHub issue for a coding agent
  • convert an issue into an implementation task
  • identify likely files and commands for an issue
  • create a task pack before assigning an agent-ready issue

What This Skill Does

It turns issue text plus the local repository into a scoped task pack containing:

  • issue summary
  • likely files
  • verification commands
  • acceptance criteria
  • risk areas
  • agent-ready prompt

Usage

From a local checkout:

python -m pip install .
issue-to-agent OWNER/REPO#123 --repo . --output ISSUE_AGENT_TASK.md

From a local issue file:

issue-to-agent examples/issue-checkout-timeout.md \
  --repo examples/mini-repo \
  --output ISSUE_AGENT_TASK.md

For a screenshot-ready report:

issue-to-agent OWNER/REPO#123 --repo . --format html --output issue-agent-task.html

For automation:

issue-to-agent OWNER/REPO#123 --repo . --format json --output ISSUE_AGENT_TASK.json

Cursor Workflow

When Cursor is the implementation agent:

  1. Generate ISSUE_AGENT_TASK.md before editing.

  2. Read the task pack and confirm the likely files match the issue.

  3. If the user asks for a reviewable artifact, generate HTML:

    issue-to-agent OWNER/REPO#123 \
      --repo . \
      --format html \
      --output issue-agent-task.html
    
  4. If another tool will consume the output, generate JSON and parse-check it:

    issue-to-agent OWNER/REPO#123 \
      --repo . \
      --format json \
      --output ISSUE_AGENT_TASK.json
    python -m json.tool ISSUE_AGENT_TASK.json >/dev/null
    
  5. Implement only the acceptance criteria after the task pack looks plausible.

Useful Cursor prompt:

Read ISSUE_AGENT_TASK.md. Use it as implementation guidance, but verify file
ranking with repository search before editing.

Copilot Workflow

When GitHub Copilot, VS Code agent mode, Copilot CLI, or Copilot cloud agent is the handoff target, ask it for one of these concrete outputs:

Use issue-to-agent with OWNER/REPO#123 and repo . Write ISSUE_AGENT_TASK.md.
Use issue-to-agent with examples/issue-checkout-timeout.md and repo examples/mini-repo.
Write ISSUE_AGENT_TASK.md, then summarize likely files, commands, risks, and acceptance criteria.
Use issue-to-agent with OWNER/REPO#123 and repo . Generate issue-agent-task.html
for maintainer review.
Use issue-to-agent with OWNER/REPO#123 and repo . Generate ISSUE_AGENT_TASK.json
for automation and verify that the JSON parses.

Agent Instructions

  1. Prefer generating the task pack before editing code.
  2. Read the generated pack and verify that the ranked files are plausible.
  3. If the ranked files look wrong, search the repository manually and mention the uncertainty.
  4. Use the suggested commands as verification guidance, not proof that the task is solved.
  5. Keep any follow-up implementation diff scoped to the acceptance criteria.

Boundaries

  • This skill does not call an LLM or hosted service.
  • This skill does not execute the implementation.
  • This skill does not publish, push, or comment on GitHub unless a workflow explicitly does so.
  • Treat auth, billing, secret handling, migrations, and destructive operations as human-review-required.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 326,984. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.