From wondraiwork
Find high-impact open source issues worth contributing to. Uses 5 search strategies (reactions, comments, help-wanted, good-first-issue, bugs) with configurable filters for stars, language, labels, and age.
How this skill is triggered — by the user, by Claude, or both
Slash command
/wondraiwork:discover [--stars=N] [--lang=LANG] [--strategy=NAME][--stars=N] [--lang=LANG] [--strategy=NAME]The summary Claude sees in its skill listing — used to decide when to auto-load this skill
You are running the WondrAIWork discovery phase (Phase 01). Your job is to help the user find
You are running the WondrAIWork discovery phase (Phase 01). Your job is to help the user find high-impact open source issues worth contributing to.
Ask the user what they're looking for. If they provided arguments, use those. Otherwise ask:
Execute the discovery script with the user's preferences:
bash ${CLAUDE_PLUGIN_ROOT}/scripts/discover.sh $ARGUMENTS
Available options:
--stars=N — Minimum repo stars (default: 1000)--lang=LANG — Language filter--label=LABEL — Additional label filter--limit=N — Results per strategy (default: 5)--max-age=DAYS — Maximum issue age (default: 90)--strategy=NAME — Single strategy (reactions|comments|help-wanted|good-first-issue|bugs)Read the methodology scoring guide:
cat ${CLAUDE_PLUGIN_ROOT}/methodology/01-discover.md
For each promising issue from the results, apply the scoring matrix:
Formula: Total = (Impact x 0.40) + (Feasibility x 0.35) + (Visibility x 0.25)
| Score | Verdict |
|---|---|
| 4.0+ | Strong target — proceed to evaluate |
| 3.0-3.9 | Worth investigating |
| 2.0-2.9 | Marginal |
| <2.0 | Skip |
Present a shortlist of 3-5 candidates in a table:
| # | Issue | Repo | Impact | Feasibility | Visibility | Total | Verdict |
|---|
For each candidate, include:
For the top candidate(s), suggest:
Use
/evaluate <owner/repo> <issue-number>to do a deep evaluation before committing to the contribution.
npx claudepluginhub kanevry/wondraiwork --plugin wondraiworkCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.