feat: workspace / team mode -- multi-machine usage tracking (shared SQLite or Postgres)#41
Open
jakduch wants to merge 2 commits into
Open
feat: workspace / team mode -- multi-machine usage tracking (shared SQLite or Postgres)#41jakduch wants to merge 2 commits into
jakduch wants to merge 2 commits into
Conversation
jakduch
force-pushed
the
feat/workspace-mode
branch
from
19e9f91 to
ae05812
Compare
jakduch
force-pushed
the
feat/workspace-mode
branch
2 times, most recently
from
a28341f to
3003d2f
Compare
Adds an opt-in shared-database mode so multiple developers can roll up Claude Code usage across a team. Two backends: - Shared SQLite (db_path on Dropbox / NFS / SMB) - Postgres (psycopg2-binary is an optional, lazy-imported dep) Schema: machine_id column added to turns + sessions tables via an idempotent migration -- existing single-machine databases upgrade in place on first scan. Dashboard: /api/data payload gains by_machine aggregation and all_machines list; a Machine filter dropdown is rendered but kept display:none until JS detects more than one machine has reported in, so single-laptop installs see no UI change. Config: ~/.claude/workspace.json (backend, machine_id, team, optional db_path, optional pg_dsn). Env vars (CLAUDE_USAGE_BACKEND, CLAUDE_USAGE_PG, CLAUDE_USAGE_MACHINE_ID, CLAUDE_USAGE_TEAM) override the file for CI / Docker workflows. Backward compat: if no workspace.json exists, behavior is identical to today's single-machine SQLite default. Tests: 22 new in tests/test_workspace.py covering config defaults, env overrides, lazy psycopg2 import, migration idempotency, pre-feature DB upgrade, machine_id stamping, by_machine payload aggregation, and UI HTML hooks. Full suite: 212 pass, 7 skipped.
…earch_wired_into_filtered_sessions passes
jakduch
force-pushed
the
feat/workspace-mode
branch
from
3003d2f to
71e6fd8
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What does this add and why do you believe it belongs in this dashboard?
This adds an opt-in workspace / team mode so the same dashboard can roll up usage from more than one machine -- either via a shared SQLite file on a network drive (Dropbox, NFS, SMB) or via a shared Postgres instance. Configuration lives in
~/.claude/workspace.json(or env varsCLAUDE_USAGE_BACKEND,CLAUDE_USAGE_PG,CLAUDE_USAGE_MACHINE_ID,CLAUDE_USAGE_TEAM); when no config exists, behavior is byte-identical to the single-machine SQLite default, so existing users see zero change.The case for a personal dashboard: a growing number of solo developers run Claude Code from a laptop and a desktop and a remote dev box, and the per-machine
~/.claude/usage.dbfragments the picture. Small teams sharing a single Anthropic plan have the same shape of problem at the next zoom level -- they need aggregate visibility into how the shared quota is being consumed before someone hits the cap mid-week. The samemachine_id-stamped schema and the same dashboard solve both cases. The UI gains a single optional "filter by machine" dropdown that stays hidden when only one machine is present, so single-machine users are never asked to think about it.Checklist
Code correctness
calcCost()calls pass 6 arguments:(model, inp, out, cache_read, cache_creation, cache_1h)-- nocalcCost()calls added or modified in this PRTests
python3 -m unittest discover -s tests -v-- all passing (212 tests, 1 skipped:psycopg2 not installed locally)python3 -m unittest tests.test_browser -v-- all passing (6 tests)tests/test_workspace.py, 28 new tests covering config defaults, env overrides, schema migration, machine_id stamping, dashboard rollups, filter UI, and lazy psycopg2 import)Scope
dashboard.py,scanner.py,cli.py,pricing.py,cowork.py,tests/) -- or I have explained below why a new file is neededworkspace.pyis added as a new module because the configuration loader, env-var precedence rules, machine-id resolution, and lazy Postgres connection helper are a self-contained concern with their own test surface; folding them intoscanner.pyordashboard.pywould have mixed transport/config concerns with scanning and rendering, and the team-mode code path needs to be importable from both.tests/test_workspace.pyfollows the existing one-test-file-per-module convention.