sync-sites

name	sync_sites
description	Sync site knowledge files with fresh Matomo baselines and events data (project)

Site Sync Skill

Update the knowledge base documents (knowledge/sites/*.md) with fresh data from Matomo and codebases.

Philosophy

This skill enriches existing documents with objective, measurable data. It does NOT:

• Generate analyses or interpretations
• Make hypotheses about user behavior
• Add commentary that could bias the agent's reasoning

The goal is to provide raw facts that the agent can use as context when answering questions. All data must be:

• Measurable: numbers from APIs, not estimates
• Verifiable: includes source links or API calls
• Dated: always shows when data was retrieved

This keeps the knowledge base clean and prevents the agent from being influenced by stale interpretations.

Document Structure

Each site document should have a clear separation between:

Auto-generated sections (managed by this skill)

• ## Traffic Baselines (YYYY) - Monthly stats tables ONLY
• ## Custom Dimensions - Dimension list from API
• ## Saved Segments - Segment list from API
• ## Event Categories - Category counts from API

Manual sections (preserved, never overwritten)

• ## Description - Site context and purpose
• ## Matomo Events - Detailed event documentation with examples
• ## Feature-Specific Insights - Analyses tied to specific features
• Any section with >10 lines of content

IMPORTANT: Put analyses and insights in SEPARATE sections, not inside Traffic Baselines. The baselines section will be fully replaced on each sync.

Usage

# Full sync (all sites, baselines + events)
python -m skills.sync_sites.scripts.sync_sites

# Baselines only (faster)
python -m skills.sync_sites.scripts.sync_sites --baselines-only

# Single site
python -m skills.sync_sites.scripts.sync_sites --site emplois

# Dry run (show what would be updated)
python -m skills.sync_sites.scripts.sync_sites --dry-run

What it syncs

Baselines (weekly)

For each site, fetches from Matomo:

• Monthly visitor stats (unique visitors, visits, daily averages)
• User type distribution (if custom dimensions exist)
• Engagement metrics (bounce rate, actions/visit, time on site)

Custom Dimensions

Fetches configured custom dimensions:

• ID, name, scope (visit or action)
• Active/inactive status

Saved Segments

Fetches saved segments from Matomo:

• Segment name and definition
• Useful for understanding pre-defined filters

Event Names

Fetches all event names directly from Matomo:

• Event name, event count, visit count
• Top 50 events shown, sorted by volume
• Data from the reference month (default: last month)

Event Discovery (manual)

When documenting events, search for these patterns in codebases:

Django templates:

• {% matomo_event "category" "action" "name" %}
• data-matomo-category=, data-matomo-action=, data-matomo-option=

Django Python code:

• matomo_event_name=, matomo_event_option=
• matomo_category=, matomo_action=

JavaScript:

• _paq.push(['trackEvent', ...])
• Matomo.trackEvent(...)

Sites

Site	Matomo ID	Doc
emplois	117	knowledge/sites/emplois.md
pilotage	146	knowledge/sites/pilotage.md
communaute	206	knowledge/sites/communaute.md
dora	211	knowledge/sites/dora.md
plateforme	212	knowledge/sites/plateforme.md
rdv-insertion	214	knowledge/sites/rdv-insertion.md
mon-recap	217	knowledge/sites/mon-recap.md
marche	136	knowledge/sites/marche.md

Weekly Schedule

Run every Monday morning:

# Cron entry
0 8 * * 1 cd /path/to/Autometa && python -m skills.sync_sites.scripts.sync_sites

Or via launchd on macOS.

Output

Updates markdown files in place, preserving:

• Header section (URL, GitHub, description)
• Custom sections you've added manually

Regenerates:

• Traffic Baselines tables
• Last sync date