Menu
Add a custom issue spotter for the paper review. Describe what to look for and this will create the spotter file.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
| name | add-spotter |
| description | Add a custom issue spotter for the paper review. Describe what to look for and this will create the spotter file. |
| argument-hint | <description of what to look for> |
| user-invocable | true |
| allowed-tools | Read, Write, Glob |
Input: $ARGUMENTS — a description of what the spotter should look for (e.g., "p-hacking", "selection bias", "missing confidence intervals").
Read the user's description from $ARGUMENTS.
Derive a short slug name from the description (e.g., "p-hacking" becomes p_hacking, "selection bias" becomes selection_bias). Use lowercase with underscores, no spaces.
Check if .review/issue_spotters/ exists. If not, create it.
Check if a spotter with this name already exists. If so, tell the user and ask if they want to overwrite.
Write a new spotter file to .review/issue_spotters/<slug>.md following this format:
# <Title>
Look for <concise description of what to check for>.
Pay special attention to:
- <specific thing to look for>
- <specific thing to look for>
- <specific thing to look for>
- <specific thing to look for>
- <specific thing to look for>
The spotter should be:
List the current spotters in .review/issue_spotters/ so the user can see what's active.
Tell the user the spotter will be used in the next /broad-sweep run.
Phase 6 — Generate the REVIEW.md report by walking the append-only review tree. Optionally create GitHub Issues.
Phase 1 — Run the chunks x issue_spotters cross-product to find candidate issues. Uses edsl for parallel LLM review.
Phase 3 — Deduplicate confirmed issues by writing a dedup.json mapping file. No files are moved or modified.
Draft author responses for each confirmed issue. Writes response.json next to each issue (append-only).
One-shot holistic referee review of the full paper. Issues are written to .review/holistic/ as append-only issue.json files.
Phase 2 — Investigate candidate issues. Writes investigation.json next to each issue.json with evidence and a confirm/reject verdict.