/sistrix-monday-recovery-check

name: sistrix-monday-recovery-check
description: 'CSV-first weekly recovery review during an active SEO recovery. No SISTRIX API key required. Compares current and previous SISTRIX keyword exports and produces a structured Monday Check report with Recovery Signal Score, per-cluster recovery stage, winner/loser neutralization analysis, money-keyword protection list, URL-level recovery table, optional GSC cross-check, optional conversion-rate validation, and one of six recommended actions (Observe / Protect / Strengthen / Investigate / Correct / Escalate) with an explicit What-Not-To-Touch guard for winning URLs. Triggers from "Monday recovery check", "SISTRIX recovery review", "weekly Monday SEO ritual", "compare this Monday SISTRIX to last Monday", "is the visibility index lagging behind ranking gains", "winner-loser neutralization check", "are my money keywords protected". Full output contract in SISTRIX_MONDAY_RECOVERY_CHECK.md; operational detail in RECOVERY_SYSTEM.md sections 4-10. CSV-only weekly review, no API access needed — for API-driven weekly tracking with a scored NDJSON history use recovery-monitor instead.'
user-invocable: true
argument-hint: '[current-csv] [previous-csv] [domain?] [money-keywords-csv?]'
allowed-tools: [Read, Write, Grep, Glob]
license: MIT
metadata:
  author: Max Schottke
  version: '0.5.4'
  category: marketing

SISTRIX Monday Recovery Check

Overview

During an active SEO recovery, every Monday the operator asks the same questions: did rankings recover, are money keywords holding, are winners and losers offsetting each other, is the visibility index lagging behind actual ranking gains. Without structure, the answer is ad-hoc and biased toward whatever the operator opens first in the SISTRIX UI.

This skill turns the Monday review into a fixed 17-section report so the same questions get answered consistently week over week. The full methodology lives in `${CLAUDE_PLUGIN_ROOT}/references/SISTRIX_MONDAY_RECOVERY_CHECK.md` (canonical spec) and `${CLAUDE_PLUGIN_ROOT}/references/RECOVERY_SYSTEM.md` (operational detail for sections 4-10). All later `SISTRIX_MONDAY_RECOVERY_CHECK.md §N` / `RECOVERY_SYSTEM.md §N` / `DECISION_ENGINE.md` mentions in this file refer to those files under `${CLAUDE_PLUGIN_ROOT}/references/`. This SKILL.md is the runnable skill that executes that workflow against the operator's CSV exports.

When to use

• A domain in an active recovery from a Google Core Update or AI Overview reshuffle and the operator wants a weekly structured pulse
• A new SISTRIX export landed and the operator wants to know "is the recovery working" without manually eyeballing 12 charts
• The operator suspects winner/loser neutralization is masking actual recovery and wants the numbers
• The operator is about to make a content or template change to a winning URL and wants a "what not to touch" check first
• An agency client deliverable: a consistent Monday report format the client can read in 5 minutes

**Don't use for:**

• Brand-new recovery diagnosis (use [[post-core-update-recovery]] for the initial diagnostic)
• A one-off audit (use [[seo-outreach-report]] for the full decision-maker PDF instead)
• Pre-recovery baselining (run [[psi-weekly-cron-baseline]] alongside instead — that is the technical-health pulse, this skill is the ranking pulse)
• Decisions about specific URL templates or code (this skill produces direction; the operator's team handles execution)

CSV-first contract

The skill does not require a SISTRIX API key. It accepts CSV exports produced from the SISTRIX UI's Keywords table. The operator exports the current week's CSV, the previous week's CSV, and optionally a money-keywords CSV.

Expected CSV shape

A SISTRIX keyword export usually carries at minimum: `Keyword`, `Position`, `URL`, `Search Volume`, `Competition`, `SERP Features`, and optionally `Previous Position`, `CPC`, and `Intent`. Column names vary by SISTRIX UI version and export type. The skill normalizes column names case-insensitively and accepts the German export form (`Keyword`, `Position`, `URL`, `Suchvolumen`, `Wettbewerb`, `SERP-Features`) as well.

If a CSV is missing one of the optional fields, the skill degrades gracefully: the section that needed that field is marked "(data not in export)" rather than producing a fabricated number.

Path arguments

/seo-rescue:sistrix-monday-recovery-check ./sistrix-exports/example-shop-2026-05-26.csv ./sistrix-exports/example-shop-2026-05-19.csv example-shop.test

If money keyword tracking is desired:

/seo-rescue:sistrix-monday-recovery-check ./sistrix-exports/current.csv ./sistrix-exports/previous.csv example-shop.test ./money-keywords.csv

The CSVs themselves are gitignored under the existing `sistrix-exports/`, `*_sistrix*.csv`, `*_keywords*.csv` rules in the repository's `.gitignore` (repo root, not shipped in the plugin package). Do not commit operator exports.

How Claude executes the workflow

This is a pure-Markdown framework skill. Claude reads the two CSVs via the Read tool, applies the methodology below, and writes the resulting report to `./output/sistrix-monday-checks/<sanitized-domain>-<YYYY-MM-DD>.md`.

Step 1 — Load and validate inputs

1. Read the current-week CSV. Detect column names. Build a normalized keyword table: `{keyword, position, url, search_volume, intent?, previous_position?}`. 2. Read the previous-week CSV. Same normalization. Build the same shape. 3. If either CSV has fewer than ~50 rows, warn the operator that the sample may be too small for reliable Winner/Loser analysis (real recovery tracking should use the full SISTRIX export, typically several hundred to several thousand keywords) and offer to continue anyway. 4. Build the keyword-keyed merge: `keywords_both = current ∩ previous`, `keywords_returned = current \ previous`, `keywords_lost = previous \ current`.

Step 2 — Visibility Index interpretation

If the operator provided current and previous V