goal-first-briefing-format

name

goal-first-briefing-format

description

How to write briefings that lead with goals + progress, not raw issue lists. Bucket every issue under a goal. The portfolio has 24+ active goals across 4 companies + Portfolio HQ; 61+ issues already linked. CEO Q4hr briefs and CoS hourly briefings BOTH lead with "this week's goal + progress through it" — issues are sub-bullets. Use this skill whenever you're about to write a briefing or status report.

Goal-First Briefing Format

Founder direction (paraphrased): "Bucket issues under goals. 1000 issues is confusing; 10 goals × 100 issues each is clear. CEO checks in on the goals and reports back."

The Paperclip data model already supports this — every issue has a goalId field, and 24 goals are active across the portfolio. What was missing was the briefing format that surfaces it.

When to use

Trigger on:

• Writing a CoS hourly briefing
• Writing a CEO Q4hr brief (whether triggered by routine or heartbeat)
• Writing a McKinsey or Paperclip Feedback strategic brief
• Writing the morning FOUNDER_ACTIONS.md
• Any status report intended for upward consumption

When NOT to use

• A doer agent's standup (those are personal-todo lists, not status reports)
• A deploy log or technical writeup (those are work artifacts, not briefings)
• An issue comment (those track per-issue progress, not goal-level)

The format (per company section)

Every per-company section in any briefing leads with this block:

### {Company name}

**This week's goal:** {goal title}
({issues in flight: X, issues done: Y, total linked: Z, M% complete})

**Progress this cycle:**
- {1-2 lines on what advanced toward the goal in the last hour/4hr}

**Issues in flight (tied to this goal):**
- {issue identifier} — {one-line status} — {agent}

**Issues in flight (NOT tied to this goal — ad hoc):**
- {issue identifier} — {one-line status} — {agent}

**Blockers / decisions needed:**
- {if any}

If the company has multiple active goals at the same priority, list 2-3 (not all 6). Pick by:

1. Critical-priority goals first
2. Goals with active in-flight issues
3. Goals nearest to deadline

Drop "achieved" goals from the briefing — they're done, surface in a one-line "Recently achieved" footer instead.

How to compute progress

For each goal, query:

# Get all issues linked to this goal
curl -s "http://127.0.0.1:3100/api/companies/<co_id>/issues" | \
  python3 -c "import sys,json; d=json.load(sys.stdin); [print(i['identifier'], i['status']) for i in (d.get('items',[]) if isinstance(d,dict) else d) if i.get('goalId')=='<goal_id>']"

Then count by status:

• done → completed
• in_progress, in_review → in flight
• todo, backlog → pending
• blocked → blocked
• cancelled → exclude from count

Progress % = done / (done + in_progress + in_review + todo + backlog + blocked) × 100

Why this matters

A briefing that opens with "33 issues in flight across the portfolio" is useless. A briefing that opens with:

Campaign Receipts — This week's goal: Ship first Tuesday newsletter by 2026-05-13 (3 in flight, 2 done, 5 pending, 40% complete) Progress this cycle: CR CTO completed Resend integration; Editorial drafted issue #1 with 4 bills.

…tells the reader exactly what matters. Same data, different framing. Founder can act on the second; they get lost in the first.

The 24 active goals (reference, as of 2026-05-03)

Each company has 6 active company-level goals + Portfolio HQ has 6:

• Concise: First $100 in direct-PDF revenue by 2026-06-15; 100 free email subscribers by 2026-06-15; First $200/mo direct MRR sustained ≥30 days by 2026-08-01; <$30/mo non-Pro spend; CEO + 4 reports work autonomously; Ship 3 growth experiments per month.
• Campaign Receipts: First Tuesday newsletter by 2026-05-13; 100 free email subscribers by 2026-06-15; First Patreon subscriber by 2026-07-01; First $5 PDF sale by 2026-07-15; <$25/mo non-Pro spend; Editorial+CTO ship newsletters weekly without founder edits.
• NT Ministry: First 10 YouTube videos by 2026-06-15; First Patreon supporter by 2026-07-15; 100 subs + 1000 cumulative views by 2026-07-15; <$50/mo non-Pro spend; Hire 4 content-production agents within 24h of approval; Self-improving content pipeline.
• HealthBrew: 100 active users with at least one lab upload by 2026-07-15; First lab upload completes parsing successfully by 2026-05-30; Biological age methodology v1 by 2026-06-15; <$15/mo non-Pro spend; Compliance Reviewer rejects 0 medical-advice violations; First feedback loop with 10 active users by 2026-07-15.
• Portfolio HQ: Daily FOUNDER_ACTIONS.md filed every morning; Hourly briefings without thrash; McKinsey 4-hr brief; Auto-promote keeps companies working; Cross-portfolio cost discipline <$50/mo; SESSION_DECISIONS stays current within 1 hour.

These rotate as they're achieved. New goals get added by CEOs (under pivot authority) or CoS.

Anti-pattern: list-of-issues briefings

Don't write briefings that look like:

Concise has 12 issues in flight: CON-3, CON-5, CON-7, CON-8...

That's a database dump, not a briefing. The reader can't tell which issues matter.

Instead always answer "what goal are we trying to hit, and how close are we?"

Related skills

• agent-autonomy-doctrine — when to act vs escalate (briefings should make escalation needs visible)
• recovery-loop-stop-pattern — banned thrash patterns that ALSO show up in goal-progress (a goal with 3 recovery-loop issues = thrash)

Severity

This is a founder-comprehension skill. If briefings don't lead with goals, founder sees noise instead of signal. The morning email digest is the most-visible application — every URGENT item should be tied to a goal it unblocks.

See BIBLE.md § 9 for the policy statement and companies/portfolio-hq/personas/chief-of-staff-hourly.md "Briefing format" section for the canonical CoS application.