Master Timeline — Citizen Analyst
Last updated: November 20, 2026
Citizen Analyst integration layer for the global master-timeline skill. Read the global skill first for the full 5-phase workflow: ~/.codex/skills/master-timeline/SKILL.md
Repository location
agent-base/timeline-repository/
manifest.json — Processing state (slug → hash, event_count, status)
consolidated.json — All events merged, sorted by date (generated)
stats.json — Summary statistics (generated)
events/ — Per-document event files (source of truth)
indexes/
by-year/
by-category/
by-subject/
KB category → extraction focus mapping
| KB category | Event focus |
|---|---|
accountability-profiles |
Official actions, scandals, investigations |
convicted-and-indicted |
Indictments, convictions, sentences, appeals |
trump-family-and-associates |
Business deals, appointments, investigations |
trump-administration |
Official actions, firings, EOs, comms |
federal-legislators |
Votes, statements, committee actions |
state-legislators |
State-level actions, elections, legal |
political-operatives |
Campaign activities, investigations, connections |
media-figures |
Statements, lawsuits, retractions |
voting |
Registration deadlines, law changes, court rulings |
legal |
Court decisions, EOs, enforcement actions |
| All others | Key developments, policy changes, incidents |
Makefile targets (from citizen-analyst/)
make timeline-extract SLUG=mark-meadows # Single document
make timeline-extract-changed # Incremental (only changed docs)
make timeline-extract-all # Full rebuild — costs $15-25 in Claude API
make timeline-reindex # Rebuild indexes only (no API calls)
make timeline-stats # Show repository statistics
Direct script usage
cd ~/Cursor/citizen-analyst
# Incremental update (recommended for routine maintenance)
python scripts/timeline_extract.py --changed
# Category-scoped update
python scripts/timeline_extract.py --category convicted-and-indicted
# Preview what would be processed
python scripts/timeline_extract.py --all --dry-run
# Stats
python scripts/timeline_extract.py --stats
Python API
Import from the scripts/ directory (or with sys.path including scripts/):
from lib.timeline_repo import TimelineRepo
repo = TimelineRepo()
# Check what's already extracted before calling Claude
if repo.is_extracted("mark-meadows"):
events = repo.events_for_slug("mark-meadows")
# Cross-document accountability research
all_acct = repo.events_by_category("convicted-and-indicted")
related = repo.events_for_subject("Steve Bannon")
j6 = repo.search_by_tag("january-6")
recent = repo.events_in_range("2025-01-01", "2025-12-31")
repo.print_events(recent, limit=10)
Workflow for new KB documents
When a new KB document is created or updated:
- Run
make timeline-extract SLUG={slug}immediately after committing the doc. - If events were extracted, the repository is updated automatically.
- The per-document
timeline-builderskill will find the events in the repository
and offer to reuse them instead of re-calling Claude.
Workflow for full rebuild
Run occasionally (e.g., monthly) or after major KB content batches:
# Preview scope
python scripts/timeline_extract.py --all --dry-run
# Build (use delay=1 to be gentle on Claude API rate limits)
python scripts/timeline_extract.py --all --delay 1
# Check results
python scripts/timeline_extract.py --stats
Expected cost: $15-25 for ~550 documents. Use --changed for routine maintenance.
Guardrails (mandatory before extraction)
- Private citizen gate: Run
patriot-private-citizen-inclusion-gatebefore extracting
events for any individual who is not a clear public official or public figure. If the gate does not pass, do not include the document in extraction scope.
- Confidence discipline: Events extracted with
confidence: lowmust be treated as
unverified and should be flagged whenever displayed. Do not publish low-confidence events to WordPress timelines without additional verification.
- Source attribution: Every event description must cite its source. Events extracted
without source attribution must be verified against the document body before use.
- Sanity check: Run
patriot-sanity-checkon any events before publishing them
to a public-facing WordPress timeline.
Integration with timeline-builder skill
The per-document timeline-builder skill creates WordPress Cool Timeline Pro stories. It should check the master timeline repository first:
repo = TimelineRepo()
existing_events = repo.events_for_slug(slug)
if existing_events:
# Offer to use existing repository events instead of re-calling Claude
# This avoids duplicate Claude API calls and ensures consistency
...
The master timeline repository stores dates in ISO 8601. The timeline_service.py _normalise_date() function converts them to MM/DD/YYYY hh:mm am for Cool Timelines Pro.
Integration with dashboard skill browser
This skill is discoverable from the Skills page in Compose, the editors’ authoring console. Invoking it from the AI Chat provides conversational guidance through the workflow.
Cross-skill references
timeline-builder— per-document WordPress timeline publishing (consumer of this repository)patriot-content-authoring— content authoring workflow (step 10: update repository)patriot-sanity-check— fact-checking extracted events before publishingaccountability-profile-verification— evidentiary standards for accountability eventspatriot-private-citizen-inclusion-gate— inclusion eligibility for individuals
