Work Hours and Overtime
wudongjieLABSCO SUMMARY
Read their privacy defaults before forming a view, because they answer the obvious objection directly.
Employee monitoring plus an AI query interface is a combination that deserves suspicion, so what they publish matters: the connection serves the derived record only and never the raw capture, no screenshots or video are stored, raw screen data is deleted during processing, keystrokes are never read, personal activity is skipped in real time, and an entry an employee redacts is erased and cannot surface here at all. Queries are permission-scoped by role. That is a considered design rather than a disclaimer. Whether the underlying product belongs in your workplace is still a management decision, not a technical one — but the redaction switch belonging to the employee is the detail that would change my answer.
What it is
ScreenJournal watches screen activity and turns it into a work record. The endpoint queries that record: what the team worked on, where the hours went, how something was done. It does not expose the capture itself.
What you get
Querying the team's work record, drafting standups, generating reports, preparing timesheets and delivering summaries. No tool list is published; queries are scoped by role permissions.
Requirements
- An MCP client your company authorises, signed in with a ScreenJournal account that belongs to an organisation; every assistant connects to the same link, https://mcp-beta.screenjournal.ai/mcp
- The desktop applications do the capture; the record is what the connection reads
- A free trial is offered; no public pricing page
Setup effort
Paste a URL, then sign in — add the endpoint to your client and sign in to ScreenJournal AI
USE CASES
01
Writing the standup
Draft it from the record rather than from memory.
02
Answering where the time went
Query hours against projects, scoped to what your role can see.
03
Filling in timesheets
Prepared from the record instead of reconstructed on Friday.
BEFORE YOU INSTALL
What to know first
Requires
- An MCP client your company authorises
- A ScreenJournal account that belongs to an organisation
INSTALLATION
Connect your client
Add the endpoint to your client and sign in to ScreenJournal AI.
Server setup
ENDPOINT URL
https://mcp-beta.screenjournal.ai/mcpRequirements
- OAuth sign-in to ScreenJournal (consent screen); account must belong to an organisation
MCP Clients
Claude Desktop, Claude.ai
Open claude.ai on a computer and sign in, then go to Settings → Connectors.
Click Add custom connector.
Paste your ScreenJournal connection link, https://mcp-beta.screenjournal.ai/mcp, and click Add.
Click Connect, and sign in to ScreenJournal when asked.
Claude Code
Add the server (--scope user makes it available in all your projects, not just the current folder):
CLAUDE CODE · TERMINAL
claude mcp add --transport http --scope user screenjournal https://mcp-beta.screenjournal.ai/mcpStart a fresh Claude Code session so it picks up the new server, then run /mcp, select screenjournal, and choose Authenticate.
Run /mcp again to confirm the server shows as connected
ChatGPT
Open chatgpt.com on a computer and sign in.
Turn on Developer mode.
Open your plugins, then click the + icon in the top right of the page.
Paste your ScreenJournal connection link, https://mcp-beta.screenjournal.ai/mcp, as the MCP server URL, choose OAuth, and sign in to ScreenJournal when asked.
