Settings & Analytics
Usage — last 30 days
Total runs
48+14.0%
vs prev 30d
Total tokens
1.2M-6.3%
vs prev 30d
Total spend
$18.40-4.2%
$ vs prev 30d
Avg run cost
$0.38
Median $0.21
Runs per day
2.4avg
Tokens / run
24.8k-12.0%
vs prev period
Spend / day
$0.61-18.0%
vs prev period
Cost attribution — this month
Monthly token budget
Token budget
· workspaceon track — will use ~35% of budget at current rateTokens1.2M / 5.0M
Cost$18.40 / $50.00USD
Breakdown1.1M segmented
- claude-opus-4-7396k
- claude-sonnet-4-6528k
- claude-haiku-4-5144k
Skills
Skills
Capabilities exposed to your agents. High-risk skills always require human confirmation regardless of toggle state.
3/5 enabled
research1
web-search
lowdefault
Brave Search API — returns top 5 results with snippets
by platform · ~800 tokens
code2
code-review
low
Reads diff and surfaces issues against project conventions
by platform · ~2,400 tokens
file-writer
medium
Writes or overwrites files on the filesystem
by platform · ~400 tokens
data1
db-query
medium
Executes read-only SQL against the connected data warehouse
by platform · ~1,200 tokens
comms1
send-email
high
Sends emails via the configured SMTP provider
by platform · ~600 tokens
Authors
Skill publishers your workspace pulls from.
platform
5 skills
Workspace settings
Notifications
Audit stream
Send audit events to the configured webhook endpoint
Cost alerts
Alert when a single run exceeds $1.00
Security
Require human approval for high-risk skills
Skills marked high-risk will always prompt for approval before running
Audit log
Writesrc/auth/session.tsby code-agent
Toolwrite_fileby orchestrator
Networkregistry.npmjs.org — jose@5.2.1by code-agent
Readsrc/auth/session.tsby code-agent
Approval.env.local — write SESSION_SECRETby avery@example.com · code-agent
Runrun_01hwf4m8kj2nqp7rby orchestrator