Kjell Tore Guttormsen 4dc868ea62 feat(ultraplan-local): v1.6.0 — /ultraresearch-local deep research command

Add /ultraresearch-local for structured research combining local codebase
analysis with external knowledge via parallel agent swarms. Produces research
briefs with triangulation, confidence ratings, and source quality assessment.

New command: /ultraresearch-local with modes --quick, --local, --external, --fg.
New agents: research-orchestrator (opus), docs-researcher, community-researcher,
security-researcher, contrarian-researcher, gemini-bridge (all sonnet).
New template: research-brief-template.md.

Integration: --research flag in /ultraplan-local accepts pre-built research
briefs (up to 3), enriches the interview and exploration phases. Planning
orchestrator cross-references brief findings during synthesis.

Design principle: Context Engineering — right information to right agent at
right time. Research briefs are structured artifacts in the pipeline:
ultraresearch → brief → ultraplan --research → plan → ultraexecute.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

2026-04-08 08:58:35 +02:00

3.3 KiB

Raw Blame History

Post Scheduling Strategy

Reference for calculating optimal posting schedule based on weekly goal and content mix.

Optimal Posting Slots

Based on weekly_goal from state file:

2x/week

Slot	Day	Time (CET)	Rationale
1	Tuesday	08:30	Peak B2B engagement window
2	Thursday	12:00	Lunch-break engagement peak

3x/week (default)

Slot	Day	Time (CET)	Rationale
1	Tuesday	08:30	Peak B2B engagement window
2	Thursday	12:00	Lunch-break engagement peak
3	Saturday	10:00	Weekend catch-up readers, less competition

4x/week

Slot	Day	Time (CET)	Rationale
1	Monday	09:00	Week-start motivation content
2	Tuesday	08:30	Peak B2B engagement window
3	Thursday	12:00	Lunch-break engagement peak
4	Saturday	10:00	Weekend catch-up readers

5x/week

Slot	Day	Time (CET)	Rationale
1	Monday	09:00	Week-start motivation content
2	Tuesday	08:30	Peak B2B engagement window
3	Wednesday	08:30	Mid-week thought leadership
4	Thursday	12:00	Lunch-break engagement peak
5	Saturday	10:00	Weekend catch-up readers

Scheduling Algorithm

When assigning dates to batch-created posts:

Start from next available optimal slot after today
Skip slots that already have queued posts (check queue.json)
If all slots this week are taken, spill into next week
Assign in slot order (earliest available first)

Slot Assignment Logic

Given: weekly_goal, today's date, existing queue entries
1. Get the slot template for this weekly_goal (tables above)
2. Find current ISO week
3. For each post to schedule:
   a. Find next available slot (date >= tomorrow, no existing queued post)
   b. Assign that date + time
   c. Mark slot as taken
4. Return list of (date, time) assignments

Format Rotation Rules

Avoid monotony by rotating formats:

No consecutive same format — If post N is "standard", post N+1 should be "carousel", "quick", "video", etc.
Suggested rotation: standard → carousel → quick → standard → video
Format weights: 50% standard, 20% carousel, 15% quick, 15% video

Pillar Balance Rules

Ensure coverage across expertise areas:

No consecutive same pillar — Enforced by topic-rotation-gate hook
No pillar >50% in a 14-day window — Also enforced by hook
Ideal distribution: Each pillar appears at least once per 2 weeks
When batching: Spread pillars evenly across the week

Time Zone Notes

All times are CET (Central European Time)
Norwegian audience peaks: 7:30-9:00 and 11:30-13:00
For international audiences, 08:30 CET catches both EU morning and US east coast pre-work
Saturday posts perform well 09:00-11:00 CET

Queue Integration

When posts are scheduled via /linkedin:batch:

Each post gets a scheduled_date and scheduled_time from this algorithm
Entry is added to assets/drafts/queue.json
Session-start hook shows today's scheduled posts
/linkedin:publish marks posts as published and updates state
/linkedin:calendar shows the full schedule view

3.3 KiB Raw Blame History