Individual skill
sqr-pipeline
End-to-end search query report (SQR) negative-keyword pipeline — pull search terms across your MCC, prep classification batches, run 3 independent classification passes for consensus, (optionally) check geo conflicts, write reviewable agree tabs, then upload approved PHRASE negatives to shared lists with two-step mutation safety. Includes a remove branch to un-negate mistakes. Auto-invoke when the user says "run SQR pipeline", "classify search queries", "pull search terms", "upload negatives", "sqr uploader", "remove negative", or "un-negate keyword".
sqr-pipeline/SKILL.md
Source commit 90f0e419f2d820789d5e50eee1237ab9fd14dfc6
What was checked
- Matches the observed repository HEAD
- Yes
- Scanner rules are current
- No
- Repository scan complete
- No
- Skill scan complete
- No
- Observed behavior
- Not checked yet
- Last checked
- Nov 11, 2026
These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.
Install this skill
Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.
An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.
Repository context
- Repository stars (not skill ratings)
- 80
- Repository last updated
- Aug 25, 2026
- License
- MIT
Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.
- ai-agents Repository topic
- ai-skills Repository topic
- claude-code Repository topic
- google-ads Repository topic
- google-ads-api Repository topic
- ppc Repository topic
- ppc-automation Repository topic
- engineering Derived category
- testing Derived category
- testing-browser Derived category
Files and excerpt
File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.
Browse files
README.mddoc · 5,086 bytesSKILL.mdcore · 11,358 bytesSKILL.mdcore · 11,358 bytesSKILL.mdcore · 11,358 bytesSKILL.mdcore · 11,358 bytesSKILL.mdcore · 11,358 bytesSKILL.mdcore · 11,358 bytesSKILL.mdcore · 11,358 bytesreferences/classify-prompt.mddoc · 5,914 bytesreferences/geo-prompt.mddoc · 5,390 bytesreferences/offbrand-keywords.txtdoc · 418 bytesscripts/mcc_search_query_report.pyruns · 15,099 bytesscripts/prep_geo_batches.pyruns · 7,117 bytesscripts/sqr_compare.pyruns · 17,956 bytesscripts/sqr_ngram_analysis.pyruns · 9,345 bytesscripts/sqr_prep.pyruns · 13,672 bytesscripts/sqr_remove_negatives.pyruns · 12,042 bytesscripts/sqr_upload_negatives.pyruns · 15,813 bytessheet-template.mddoc · 5,833 bytes
Read excerpt
# SQR Pipeline — Negative Keyword Operator
One operator owns the search-query-report negative-keyword workflow end to end:
pull → prep → 3-run classify → (optional geo) → consensus → human review → upload,
with a maintenance branch to remove a negative that was added by mistake.
The classification step is **LLM-in-the-loop** — Claude reads each batch and
classifies it three independent times. That is the whole point: three independent
opinions per query, with consensus (3-of-3 unanimous / 2-of-3 majority) filtering
out one-off misclassifications before anything touches a live account.
This is a single generic vertical. There is no profile switching and no
account-specific branching — point it at *your* MCC, *your* Google Sheet, and
*your* competitor list (`references/offbrand-keywords.txt`) and it runs.