Create Issue
Create a new issue file from a test262 failure pattern. Includes smoke test, sample extraction, and proper frontmatter.
Get the complete skill folder, including its scripts and reference files.
When to use this skill
Use Create Issue when an AI agent needs a reusable SKILL.md workflow for this job: Create a new issue file from a test262 failure pattern. Includes smoke test, sample extraction, and proper frontmatter.
When not to use it
Skip Create Issue when the task is outside the coding category, or when a more specific skill in this directory already covers the same workflow with clearer triggers.
How to install
Where will you use it?
mkdir -p ~/.claude/skills/create-issue # Save SKILL.md to ~/.claude/skills/create-issue/SKILL.md
- Claude Code personal install: place the skill and its bundled files at ~/.claude/skills/create-issue/SKILL.md. For a shared project, commit the folder at .claude/skills/create-issue/ instead.
- Current Claude Code detects SKILL.md changes in watched directories during the session. If you just created a top-level skills directory, run /reload-skills. Check that the skill is listed before testing its trigger.
- For Claude chat or Cowork in the Desktop app, manage uploaded skills in Customize > Skills; copying files into a local Claude Code folder is not a Desktop install. Other agents may use different install locations.
Full install guide for Claude, Cursor, and Codex →
Try a starter prompt
After installing the skill, paste this starter prompt into your agent to verify it triggers on a small task.
Use the Create Issue skill for this task: Use Create Issue when this documented workflow matches the task. Confirm any required inputs or files before making changes.
What supports this recommendation?
68/100 provisional content score. This measures documentation, practical detail and source signals. It is not a measured task success rate or a security certification.
- Documentation
- 72/100
- Practical detail
- 68/100
- Source evidence
- 58/100
- Source signals
- 72/100
Model compatibility evidence
- claudeinferred
The skill text mentions Claude or a closely associated term. This is keyword-derived compatibility; no hands-on model test is recorded.
- chatgptuntested
No model-specific signal or recorded compatibility test was found.
- geminiuntested
No model-specific signal or recorded compatibility test was found.
- copilotuntested
No model-specific signal or recorded compatibility test was found.
- llamauntested
No model-specific signal or recorded compatibility test was found.
- perplexityuntested
No model-specific signal or recorded compatibility test was found.
- mistraluntested
No model-specific signal or recorded compatibility test was found.
- grokuntested
No model-specific signal or recorded compatibility test was found.
Inferred compatibility is based on content signals. Check the app, its available tools and setup requirements before running a skill.
What this skill does
# Create Issue from Failure Pattern
## Input
Describe the failure pattern: error message, test category, approximate count.
## Steps
1. Reserve the next issue id **atomically** (#2531) — never hand-pick a number (a parallel PR may grab the same one; the dup only fails in the merge_group and wedges the queue): ```bash NEW=$(node scripts/claim-issue.mjs --allocate) # prints the reserved id echo "$NEW" ``` This reserves the next id unique against `origin/main` ∪ every open PR's added issue files ∪ ids already reserved on the orphan ref, with first-push-wins atomicity. Use `$NEW` as `{N}` below. (The CI gate `check:issue-ids:against-main` rejects any PR that introduces a main-colliding id, so this is enforced, not just advised.)
2. Find sample test files matching the pattern: ```bash grep -l "<error pattern>" /workspace/benchmarks/results/test262-results.jsonl | head -5 ``` Or search the JSONL: ```bash python3 -c " import json matches = [] with open('benchmarks/results/test262-results.jsonl') as f: for line in f: r = json.loads(line) if '<pattern>' in r.get('error',''): matches.append(r['file']) print(f'Found {len(matches)} matching tests') for m in matches[:5]: print(f' {m}') " ```
3. For each sample, extract the relevant source lines from the test file.
4. Write the issue file: ```markdown --- id: {N} title: "{description} ({count} tests)" created: YYYY-MM-DD updated: YYYY-MM-DD priority: {high|medium|low} feasibility: {easy|medium|hard} depends_on: [] goal: core-semantics es_edition: {es5|es2015|es2016|es2017|es2018|es2019|es2020|es2021|es2022|es2023|es2024|multi|n/a} language_feature: {normalized-feature-slug} task_type: {bug|feature|test|refactor|planning} ---
# #{N} -- {title}
## Problem
{count} tests fail with {error pattern}. {1-2 sentence description}.
### Sample files with exact errors and source
**1. {test name}** File: `{path}` Error: `{exact error}` ```js // Relevant source lines ``` Root cause: {why the compiler produces wrong output}
## ECMAScript spec reference
{Link to the relevant spec section(s) at https://tc39.es/ecma262/} Example: [§7.1.1 ToPrimitive](https://tc39.es/ecma262/#sec-toprimitive)
If multiple sections apply, list each with a one-line note on which step or sub-algorithm is relevant to this issue.
## Root cause in compiler
{Which codegen function/file produces this pattern}
## Suggested fix
{High-level approach — specific enough for architect to spec}
## Acceptance criteria
- {specific test files that must pass} - >={target} of {total} tests fixed ```
Required metadata rules: - `created` is the first known creation date and should not change later - `updated` must be bumped whenever the issue meaningfully changes - `es_edition` should capture the relevant language edition, or `multi` / `n/a` - `language_feature` should be a stable machine-readable feature or subsystem slug - `task_type` must be one of `bugfix`, `feature`, `test`, `refactor`, `planning`
5. Add to dependency graph and backlog.
Intended uses
- Use Create Issue when this documented workflow matches the task.
Implementation notes
- Review the source instructions and adapt inputs before running the workflow.
Try it, then tell us how it went
A copied prompt does not tell us whether the task worked. Report your result after checking the output. We collect your tool and outcome, never your files or prompt.