ops-cadence

name	Ops Cadence
description	Run the weekly FrankX operations loop — triage the inquiry inbox, advance the CRM pipeline, sweep open PRs across repos, and surface the content plan. Use for "/ops", "run the ops loop", "weekly ops", "triage inquiries", "what needs my attention", or any business-operations check-in.
version	0.1.0
triggers	["/ops","run the ops loop","weekly ops","triage inquiries","what needs my attention","ops check-in"]
inputs	{"required":[],"optional":["mode","since"]}
outputs	["A prioritized ops readout (inbox · pipeline · PRs · content)",{"Optional":"drafted replies, advanced Notion stages, PR actions"}]
dependencies	["frankx-brand"]
mcp	["notion-mcp","github-mcp","slack-mcp"]

Ops Cadence

The execution layer of the FrankX Operations System. One command surfaces everything that needs Frank's attention across the business and drives each item one step forward. Pairs with the capture/acknowledge/track layers built into frankx.ai (/api/intake → Notion "Inquiries" CRM). See docs/ops/OPERATIONS_SYSTEM.md in the website repo for the full architecture.

When to run

• Weekly (Monday) — the full loop, as a planning ritual.
• On demand — "what needs my attention?" before a focus block.
• After a launch — when inbound spikes and the pipeline needs triage.

The loop

Four passes, each ending in a concrete action — not just a report.

1. Inbox — triage new inquiries

Source: Notion Inquiries (CRM) database (or /admin/intake if Notion isn't wired yet). Pull every row at stage New.

For each:

• Classify urgency (commercial intent + named company + specific ask = hot).
• Draft a reply in Frank's voice (technical-authority register — see the frankx-brand skill). Lead with the answer, propose the next step (call / scope / no-fit).
• Advance the Notion Stage: New → Triaged.

Output: a ranked list — hot leads first, each with a ready-to-send draft.

2. Pipeline — advance the CRM

Pull rows at Triaged, Call booked, Proposal.

For each, name the single next action and who owns it:

• Triaged with no reply sent → send the draft.
• Call booked with date passed → log outcome, move to Proposal or Lost.
• Proposal aging > 7 days → nudge or mark Lost.

Flag anything stuck > 14 days. Stale pipeline is the silent killer.

3. PRs — sweep open work

Across frankxai/frankx.ai-vercel-website and frankxai/agentic-creator-os:

• List open PRs with CI status + review state.
• Green + reviewed → flag for merge.
• Red CI → diagnose (one line) + propose the fix.
• Stale draft (> 7 days, no activity) → decide: finish, or close and capture the idea in BRANCH_AUDIT.md.

Output: a merge queue and a fix queue.

4. Content — surface the plan

• What shipped since last loop (blog, builds, social)?
• What's queued / half-built (draft PRs, content/builds/* at status: wip)?
• One recommended next piece, tied to a real build (feed the /build-log flywheel).

Modes

Mode	Scope
full (default)	All four passes.
inbox	Just triage new inquiries + draft replies.
pipeline	Just advance CRM stages.
prs	Just the PR sweep.

Output shape

## Ops Readout — <date>

### 🔥 Hot (act today)
- <Name, Company> · <intent> — <one-line ask> → [draft ready]

### 📥 Inbox (N new)
...

### 📊 Pipeline (N active)
- Stuck > 14d: ...
- Next actions: ...

### 🔀 PRs
- Merge queue: ...
- Fix queue: ...

### ✍️ Content
- Shipped: ... | Queued: ... | Recommend next: ...

Guardrails

• Never auto-send a reply or advance a Won/Lost stage without Frank's confirmation. Draft and propose; Frank approves.
• Voice: technical authority. Lead with the answer, link the artifact, let the work speak. No hype, no guru language.
• Honesty: if the pipeline is empty or PRs are all clean, say so plainly. An ops loop that invents busywork is worse than none.