Every session together
Which session is working on what and which one is waiting for an answer, live.
Online · One screen, every session
See every AI session running at once, the tasks they are on and the services behind them, side by side. What is running, what is waiting, what broke: one look tells you.
Good afternoon · first session
Welcome,
Admin
Welcome back, have a good day.
Tip of the day medium
Long sessions fill the context; when a task is done, summarise with /compact and carry on.
System health
100/100
K 100 · Ö 100 · H 100
Optimization
97/100
182 artifacts · Artifact quality very good
● All factors healthy, nothing pending
▸ All factors (12)
Optimization actions
The score shows on the System Health card
Ecosystem inventory
2 min96
Skill
44
Command
34
Agent
18
Rule
61
Hook
312
Memory
188
Instinct
12
PM2
7
MCP
3
Cron
Fleet
9
Active Claude sessions
Sessions[ Today ] [ Week ] [ 30 D ]
402
Week · Total 1,214
Token usage[ Today ] [ Week ] [ 30 D ]
19.8M
CLI 19.8M · HCD 0
Quick actions
Recent activity (5)
All rolesAll →| Date | User | Role | Action | Source | Detail | IP |
|---|---|---|---|---|---|---|
| 12.10.2026 14:53 | Admin | SUPERADMIN | signed in | auth | Session opened | 203.0.113.24 |
| 12.10.2026 13:20 | Admin | SUPERADMIN | created a task | tasks | Prepare release notes | 203.0.113.25 |
| 12.10.2026 11:46 | Admin | SUPERADMIN | updated a setting | settings | Notification hours | 203.0.113.26 |
| 12.10.2026 09:05 | Admin | SUPERADMIN | signed in | auth | Session opened | 203.0.113.27 |
| 11.10.2026 18:32 | Admin | SUPERADMIN | added memory | memory | Pre-release checks | 203.0.113.28 |
Active Claude Code sessions
Total
9
9 active, 0 idle
Active
9
active in the last 60s
Idle
0
all active
Tagged
0
9 without a role
Role [ All 8 ] [ No role 8 ]
Task tracking and management
Total
6
Waiting
2
In progress
2
Done
2
Live state of the persistent lead session (read only)
What got done today?
Three things closed: the landing page mobile layout is live, the backup was verified and two tickets were answered. One decision is waiting for you: the weekly backup check.
Leave the decision for tomorrow.
Done, it is on the agenda for 09:00 tomorrow.
Health: all sources fine
38 sources watched · Details
PM2, cron and system status
PM2 processes
12
Scheduled jobs
3
Memory
7.4 G
/ 16 G (46%)
CPU load (1m / 5m / 15m)
1.2 / 0.9 / 0.8
8 core
api-gatewayrunning
PID 1201 · CPU 0.4% · MEM 48.2 MB · UP 2g 4sa
worker-queuerunning
PID 1202 · CPU 1.1% · MEM 61.7 MB · UP 2g 4sa
mail-bridgerunning
PID 1203 · CPU 0.1% · MEM 22.5 MB · UP 2g 4sa
schedulerrunning
PID 1204 · CPU 0.2% · MEM 18.9 MB · UP 2g 4sa
docs-indexrunning
PID 1205 · CPU 0.3% · MEM 74.0 MB · UP 19sa 12dk
voice-sttstopped
PID 0 · CPU 0% · MEM 0 B · UP 0dk
Every session, queue and service on one screen
Sessions
9
Queue
4
Services
12
Alerts
1
web-lider
Landing page copy · 12 min
oyun-lider
Balance run · 41 min
bakım kuyruğu
4 jobs waiting
yedekleme
Last run 03:10, ok
Talk to a model directly, with history and templates
Chats
38
This week
7
Templates
11
Models
4
Release note draft
today 10:24 · 14 messages
SQL query review
yesterday · 6 messages
Campaign copy
2 days ago · 9 messages
Messages and questions sessions left for you
Unread
3
Questions
1
Today
8
Archived
214
web-lider
Release ready, waiting for approval
oyun-lider
Measurement table added
destek-lider
3 tickets answered
Claude Code team management
Teams
2
Members
7
Open tasks
5
Done
19
site-yenileme
4 members · lead web-lider
oyun-dengesi
3 members · lead oyun-lider
Claude Code agent library and live run tracking
Agents
34
Running
2
Today
16
Failed
0
code-reviewer
Code review · 3 runs today
tdd-guide
Tests first · 2 runs
docs-lookup
Library docs · 5 runs
security-engineer
Security scan · 1 run
Autonomous work layer: jobs, runs, pending decisions and permission areas
Jobs
27
Runs today
64
Pending decisions
2
Failed
1
site-check
Daily · last run 07:00 ok
memory-trim
Daily · last run 04:30 ok
docs-sync
Hourly · retrying
backup-verify
Weekly · waiting for a decision
External AI channels: health, quota, queue, recent jobs and outputs awaiting review. Read only.
Channels
6
Healthy
5
Queue
2
Awaiting review
1
codex
Quota 38% · last job 9 min ago
local model
GPU idle · queue empty
deep-research
1 report awaiting review
Bank accounts, cards, gold and currency; updated from notifications on their own.
Net worth
₺248.300
Card debt
₺12.450
Spent this month
₺18.920
Accounts
5
Sample Bank checking
Last movement today 09:12
Sample card
Due 14 November
Gold (grams)
Price feed 5 min ago
Accuracy of the local decision models per flow, their decisions and your corrections.
Accuracy
%94
Decisions today
31
Corrections
2
Flows
8
mail labelling
97% · 212 decisions
notification priority
91% · 88 decisions
reply needed
86% · 40 decisions · 2 corrections
Collected postings, shortlist, application status and source health.
Postings
146
Shortlist
9
Applied
3
Sources
7
Senior Backend Developer
Sample Tech · remote · score 86
Platform Engineer
Sample Software · hybrid · score 79
Game projects: roadmap, test worlds and release status
Projects
2
Open items
23
Test worlds
1
Last build
0.9.4
Sample strategy game
Testing · 14 open items
Sample card game
In development · 9 open items
CI, quality gates and next step for each project
Projects
6
CI green
5
Open PRs
2
Red
1
Sample SaaS
CI green · 2 open PRs
Personal site
CI green · deployed 2 hours ago
Mobile app
Lint red
Past Claude Code sessions and their summaries
Today
18
Week
402
30 days
1.214
Avg length
38 dk
web-lider
Landing page layout · 1 h 12 min
oyun-lider
Balance run · 48 min
destek-lider
Ticket replies · 22 min
Token use by model, project and day
Today
3.4M
Week
19.8M
Cache hits
%81
Top model
opus
opus
12.1M · 61%
sonnet
6.2M · 31%
haiku
1.5M · 8%
Persistent memory files
Files
312
Feedback
168
Project
61
Reference
83
Short commit messages
feedback · 3 days ago
Pre-release checks
feedback · 1 week ago
Sample SaaS architecture
project · 2 weeks ago
Learned patterns and rules
Instincts
188
This week
6
Promoted
14
Archived
22
Reproduce the bug first
debugging · 9 observations
One probe is not a fact
verification · 14 observations
Skills, commands, agents, rules, hooks
Skill
96
Commands
44
Agents
34
Hook
61
review-it
command · used today
brainstorming
skill · 12 uses this week
secret-scan
hook · before commit
Service connections and tool integrations per project
Connected
41
Attention
2
Off
5
Categories
9
GitHub
Repos, PRs, CI
Google Takvim
Read only
Telegram
Bot token needs renewal
BM25, vector and rerank search over a local markdown index.
Documents
4.812
Collections
17
Hit score
0.91
Last index
06:00
docs
1,204 documents · fresh
memory
312 documents · fresh
notes
2,088 documents · 1 day ago
Notifications from services and sessions
Unread
4
Warnings
1
Errors
0
Today
23
Deploy finished
Personal site · 2 hours ago
Disk at 80%
System · 5 hours ago
Backup verified
Maintenance · today 04:31
General, dashboard, security and system configuration
Users
3
API
2
2FA
on
Session length
8 sa
Two-step verification
Authenticator code or fingerprint
Remote access policy
Risky actions only from the home network
API keys
With scope, expiry and IP list
HCD REST endpoints, examples and permission scopes
Endpoints
186
Scopes
24
Version
v1
OpenAPI
3.1
GET /api/v1/tasks
Lists tasks · tasks:read
POST /api/v1/tasks
Creates a task · tasks:write
GET /api/v1/fleet
Active sessions · fleet:read
HCD
Built to keep track of dozens of jobs running all day. Everything in one place, in one language.
Which session is working on what and which one is waiting for an answer, live.
Give Jarvis work by typing or speaking; it hands each job to the right session.
Open work, its steps and who is on it sit in one list.
When a service stops or a job gets stuck, the panel tells you before you go looking.
Search notes, documents and past decisions in plain language.
Sign-in asks for a password and a phone code, every action is recorded. Risky actions only work from the home network.
HCD is a panel I built for my own work. If you want something similar for your team, write and we will talk.
hello@hakantopcu.dev