Docs
Aetower Tab-by-Tab Guide
Current app workspaces, nested sections, and when to use each page.
This guide describes the app as it is currently organized: eight top-level workspaces with nested sections for Activity, Storage, Repos, Agents, System, and Settings.
Updated: 2026-07-14 Applies to: Developer Preview 0.9.0
1. Monitor
What it does
Monitor is the default triage page. It groups running processes into user-meaningful entities, ranks them by friction, and shows the host pressure signals that explain why the Mac feels slow, hot, loud, battery hungry, or memory pressured.
Use it for
- finding the highest-friction app, daemon, terminal session, or AI runtime
- comparing CPU, resident memory, wakeups, process count, disk, network, and energy burden in one list
- opening entity detail to inspect process membership, origin, trends, and recommendations
- previewing guarded process actions when operator actions are enabled
Why it exists
Activity Monitor exposes process rows. Aetower's Monitor exposes the operational unit the user actually recognizes: the app, service, session, or runtime causing the burden.
2. Activity
Activity is the time-domain workspace. It combines what used to be separate History and Timeline concepts into one navigable area.
Overview
Shows recent changes, event pressure, history coverage, and load status.
Use it when you need a quick answer to "did anything meaningful change recently?"
History
Shows persisted snapshots, historical trends, recurring entities, and before/after comparisons.
Use it when you need to answer "what was running yesterday at 2 PM?" or verify whether a fix actually reduced friction later.
Timeline
Shows notable events in chronological order: launches, exits, restart loops, host pressure changes, thermal changes, alerts, and AI-session markers.
Use it when you need the story of an incident instead of a static snapshot.
Storage
Shows the health of Aetower's own history store: DB/WAL size, retention window, quarantine state, and maintenance status.
Use it when a history page is stale, incomplete, or heavy to load.
3. Storage
Storage is the local reclaim and disk-pressure workspace. It separates quick operator decisions from deep investigation.
Scan Controls
The top bar exposes the simple scan profiles:
- Quick scan: fast changed-state scan for day-to-day checks
- Complete scan: full normal scan for operator review
- Forensic scan: most complete profile, used when accuracy matters more than speed
- Custom scan: optional root/depth/limit controls for narrow investigations
While a scan runs, Storage should show visible progress/state so the user knows work is happening.
Reclaim
Reclaim is for immediate space-saving decisions.
It includes:
- disk pressure and reclaim summary
- quick wins
- duplicate-file quick win
- old screenshots, defaulting to "older than 7 days"
- file/folder review where relevant
- stage cleanup, move to Trash, and copy cleanup plan actions
Reclaim should not duplicate the full Explore inventory. Its job is to surface high-confidence actions and next reviews.
Similar
Similar is the full redundancy review page.
It groups:
- exact duplicates
- similar images
- similar documents/text
- similar videos
- similar binaries
- other redundancy
Use it when the question is "which files are the same or close enough that I should review them together?"
Explore
Explore is for deep storage investigation.
It contains:
- Browse: indexed, paged table
- Optimize: large files, app footprints, System Data, and investigation leads
- Cold: old candidates ranked by age and reclaim potential
- Raw: retained artifacts for deep inspection
- visual modes: full disk, treemap, and table
Use it when Reclaim is too summarized and you need to inspect the actual storage shape.
Audit
Audit is the local cleanup record.
It shows staged, moved, blocked, overridden, already-reclaimed, and failed cleanup actions so storage work stays explainable and reversible.
Insights
Insights covers growth over time, source/volume coverage, budget guardrails, and raw diagnostics.
Use it when scan output looks surprising or stale.
4. Repos
Repos is the local workspace truth layer. It starts from cached repository inventory, checks fingerprints for freshness, and overlays Git, storage, agent, and cost context.
Overview
Shows the repository list with status, Git state, inventory freshness, attention signals, AI usage, project links, and primary actions.
Sorts include:
- Attention
- Size
- Growth
- Artifacts
- AI spend
- Name
Attention
Focuses on repos needing review: stale fingerprints, dirty worktrees, duplicate clones, missing agent contracts, growing storage, review items, or live agent activity.
Repo Detail
Each selected repo can expose:
- Actions: recommended operator actions and prompts
- Storage: repo artifact footprint and cleanup context
- Contracts:
AGENTS.mdand.agents/*.yamlreadiness - Scorecard: OpenSSF/GitHub supply-chain checks when available
- Git: branch, HEAD, remote, status, duplicate clone context
- Live: linked sessions and runtime activity
Most actions should reveal folders, copy briefs, or prepare local contract kits instead of mutating the repo blindly.
5. Projects
Projects links repositories to provider context.
It includes:
- GitHub connector
- Cloudflare connector
- project records created from local repositories
- provider status overlays such as pull requests, workflow runs, checks, and Cloudflare deployments
Use it when the question is "what does this repo map to outside the local filesystem?"
6. Agents
Agents is the AI-workload workspace.
Chau7
Shows Chau7 session context when Chau7 is available: terminal sessions, tabs, repositories, branches, active apps, status, and session materialization.
AI Agents
Shows local AI tools and runtimes, including coding agents and model runtimes. It surfaces burden leaders, session energy/cost context, memory pressure, GPU inference labels, and source/confidence badges where attribution is heuristic.
Use it when the question is "what is this AI work costing the machine or this repo?"
7. System
System is for machine, startup, diagnostics, and local-network state.
Sensors
Shows hardware and host health: thermals, fans, power, storage, battery, Bluetooth, and per-core load.
Startup
Shows startup and persistence inventory.
Sections:
- Summary
- Attention
- Active Now
- All Items
- Locations
Use it when you need to know what starts automatically, what is active now, and what deserves review.
Diagnostics
Shows Aetower's own health.
Sections:
- Health
- Crash / Reboot
- MCP
- UI Payloads
- Memory
- Event Stream
Use it before filing an issue, debugging an adapter, or exporting a support bundle.
Fleet
Shows nearby Macs running Aetower with Fleet enabled. Fleet is opt-in and meant for trusted local networks.
Use it to compare local Mac health without a cloud account.
8. Settings
Settings controls the behavior and trust posture of the app.
Sections:
- Setup: first-run checklist and MCP consent card
- General: UI and operator-safe behavior
- Collection: cadence, adaptive behavior, and sampling depth
- Repositories: repository roots and repo-related behavior
- Integrations: Chau7, Docker, Chromium, provider endpoints
- AI Clients: MCP registration and CLI status
- Notifications: alert thresholds and channels
- Automation: local automation rules
- Privacy: outbound data status and privacy controls
- Updates: Sparkle update status
- Advanced: reset/support/advanced local data operations
Use Settings whenever enabling something that can change collection depth, write client config, register MCP, export data, or reset local state.
Aetower is a free early-alpha download for macOS 14+ (Apple silicon).
Download for macOS