CEO Dashboard
The systems you use to run your work, all in one place.
My operating systems
Open the workspace you need.
Tools
Projects
Quick Capture
Dump the thought here first. Organize it when you have a minute.
Your capture inbox
Move notes along as you sort and delegate them.
Labor costs
Hubstaff hours and pay costs in Philippine pesos and US dollars.
Gusto
Connect the CEO Hub to Gusto through Gusto's official sign-in.
Connect Gusto
Approve payroll and contractor read access. The connection stays separate from Hubstaff.
Ask Hub AI
Use the same OpenRouter-powered assistant across your dashboard, build library, tools, and operating work.
Your conversation stays in this browser session. OpenRouter is called through the protected CEO Hub server; the API key is never sent to the browser.
Hubstaff screenshot
Upload the timesheet approval screen. The assistant reads each visible Total Worked value.
Processed for this request. The screenshot is not saved in Owner Hub.
Gusto hours
Hubstaff time stays visible. Any changes appear separately before the final number.
Tools & References · Internal Ops
Tools & References
Open a resource to view its details.
Tools & References · Internal Ops
AI Coding Models & Usage
These are consumer-level comparisons based on how I should use each model, not claims that the models are technically equivalent.
| OpenAI | Claude | Best use |
|---|---|---|
| Astra | Fable | Highest-end capability, hardest and most complex work |
| Sol | Opus | Heavy reasoning, complex coding, difficult problems |
| Terra | Sonnet | Strong everyday workhorse |
| Luna | Haiku | Fast, economical, routine and basic work |
| Codex Spark | N/A | Ultra-fast specialized coding, use separate allotment when available |
Default
- Codex: start with Spark when appropriate, otherwise Terra
- Claude: start with Sonnet
Escalate only when needed
- Terra/Sonnet struggles: move to Sol/Opus
- Sol/Opus struggles, or the task truly requires maximum capability: move to Astra/Fable
Use lower models for
- Routine coding, UI changes, copy and text changes, repetitive implementation
- Browser clicking and navigation, data entry, and copying information between systems
- Following an established implementation plan and making simple fixes
Use higher models for
- Architecture decisions, complex debugging, major migrations, and security or authentication
- Difficult database decisions, complex integrations, and problems where the correct approach is unclear
- Reviewing high-risk or consequential changes
Important: Sol High is Sol using a higher reasoning level, not a separate model tier. Do not automatically use High or maximum reasoning. Long or tedious does not necessarily mean difficult. Keep Fast mode off unless speed is worth the additional usage. Avoid unnecessarily long threads, start fresh when moving to a substantially different feature or problem. Put permanent project knowledge in repository documentation instead of relying on giant conversation threads. Use the cheapest or lowest model that can reliably complete the task.
Tools & References · Internal Ops
Lovable Cost & Credit Tracker
A clean reference for what Lovable credits mean in dollars and where the major OS builds used them.
Lovable Credit Reality Check
Conservative cost gauge: 1 credit = $0.30.
| Credits | Conservative cost |
|---|
Credit Usage Tracker
Major OS projects · estimated cost = credits × $0.30.
| Project | Credits Used | Est. Cost | % of Usage |
|---|
Lovable Spend Breakdown
Actual cash paid to Lovable.
| Type | Amount Paid |
|---|---|
| Pro 2 subscriptions | $300 |
| Credit top-ups | $345 |
| Total actual cash paid | $645 |
Actual Lovable spend
The conservative value is not necessarily what you paid for those credits. It deliberately uses the higher top-off rate as a grounding gauge.
Keep this in mind
About 100 credits left is roughly $30 of conservative build capacity. A 200-credit top-off is about $60. A 500-credit top-off is about $150.
What we're building
Use the row dropdowns to change type or status