Log in Track your first session

Your team’s AI sessions,
in one library.

Questions blueplane answers from your library: Where is our token spend going? Which model for which task? What did we ship last week? Who is using this and who is not?

API & MCP let you ask the questions specific to your team or workflow.

Works with +more

Every session your team ran last week is gone. Blueplane means you never lose a session ever again. Your AI session library only gets more valuable over time.

Complete data stored for every session.

Prompts, responses, tool calls, skill invocations, and everything else is all recorded.

YouTeam MK
AI sessions
All sessions1,284 By project By work type By week By person
Analytics & learning
Dashboard Autonomy Level Resources Chat
All sessions· 1,284 sessions · newest first Search sessions
projectwork typetool sort: newest
Today
Add rate limiting to the public search API
search-svc·feature·Cursor·2:41pm·18 min·22 tools
DO
Migrate billing events to the new ledger schema
billing·migration·Codex·2:36pm·44 min·61 tools
RS
Why does checkout retry twice on a 502?
checkout-svc·debugging·Claude Code·2:30pm·26 min·38 tools
MK
Draft the Q3 incident review from on-call notes
ops-docs·writing·Claude Cowork·2:24pm·31 min·12 tools
JV
Scaffold settings page with shared form hooks
web-app·feature·OpenCode·2:19pm·52 min·74 tools
AT
Extend audit logs with before/after payloads
platform-api·migration·Claude Code·1:58pm·46 min·23 tools
MK
Flaky e2e: cart badge count off by one
web-app·debugging·Cursor·1:12pm·38 min·29 tools
SP
Spike: DuckDB vs Postgres for the events rollup
data-pipeline·exploration·Codex·11:40am·41 min·17 tools
NK
Backfill derived titles from session summaries
platform-api·refactoring·Claude Code·10:05am·1h 04m·55 tools
DO
Yesterday
Tune checkout rate limits
checkout-svc·refactoring·Cursor·6:20pm·38 min·19 tools
RS
Make the CLI installer idempotent
cli·feature·OpenCode·4:02pm·41 min·33 tools
AT
Illustrative data

Team library across all platforms.

Don’t get locked into a single AI platform. Own your data across all platforms your team uses today and in the future.

Ask it anything.

API & MCP mean you can ask the questions that are specific to your team and workflows.

Where is our token spend going?
blueplaneMCP · Claude Code

Three repos drove 62% of last week’s tokens. Most of it came from debugging sessions that ran past two hours.

checkout-svc31%
data-pipeline18%
web-app13%
everything else38%
From 412 sessions this week
Which model for which task?
blueplaneDashboard chat

Migrations land on the first try far more often on the larger model. For test scaffolding, the smaller one does just as well for less.

Work typeBest first-try modelFirst try
migrationsclaude-opus-571%
test scaffoldingclaude-sonnet-564%
debugginggpt-5.2-codex58%
From 388 sessions · last 30 days
What did we ship last week?
blueplaneAPI

9 pull requests merged out of 57 sessions. The biggest was the audit-log payload migration.

#1184Extend audit logs with before/after payloads4 sessions
#1179Rate limiting for the public search API3 sessions
#1171Billing events on the new ledger6 sessions
From 57 sessions · 7 repos
Who is using this and who is not?
blueplaneMCP · Cursor

6 of 8 engineers ran sessions on most days. Two haven’t started one in two weeks, both new to the mobile team.

MKMaya K.48
DODev O.41
NKNoor K.0
LBLena B.0
From 1,284 sessions · last 14 days
Ask your library anything
Illustrative data

Rich data is stored in how the work gets done.

Work happens 10x faster, and you need tools that make that work visible and legible.

Extend audit logs with before/after payloads

MK
platform-api·Claude Code·Sep 12·1:58pm–2:44pm·46 minCollapse all
0:00◆
▾
Turn 1 · claude-opus-5
3m 12s · 9 tools · 1 subagent
0:00❮
User Prompt Extend audit_logs with before and after payloads as jsonb, plus a retention window. Migration and the writer.
0:04≡
▾
searched the repo · read files (9 tools) · 1 subagent
✻
The writer already calls scrub() before insert, so both payloads should go through that same path.
0:06⌁
Grep — audit_log
0.4s
0:07±
writer.py (read)
0:09⑂
▾
Launched 1 subagent in parallel · 1 completed
0:09⑂Explore · 12 tools · 16.2k tokens · 18 messages
✻Find every reader of audit_logs so the new columns don’t break a downstream query.
1:01⑂completed after 52s
❯Three readers: admin export, the compliance report, and the nightly rollup. None of them select every column.
1:14❯
Agent Response One migration adds before_state, after_state and retention_until, and both payloads route through scrub().
2m 41s idle
3:55◆
▾
Turn 2
11m 06s · 14 tools
3:55❮
User Prompt Go ahead. Write the tests first.
3:58≡
▾
invoked skill · edited files · ran commands +116 −13 (14 tools)
⌁
Skill — tdd
0.2s
4:10±
test_audit_payloads.py +64 −0 · write
4:31±
▾
0042_audit_log_payloads.sql +48 −0 · edit
1+ALTER TABLE audit_logs
2+ ADD COLUMN before_state jsonb,
3+ ADD COLUMN after_state jsonb,
4+ ADD COLUMN retention_until timestamptz;
5:02±
writer.py +3 −12 · edit
6:40⌁
Bash — pytest tests/audit -q
2.4s · error
6:41✻
test_retention_is_tz_aware fails: retention_until is built from a naive datetime.
7:15±
writer.py +1 −1 · edit
7:22⌁
Bash — pytest tests/audit -q
2.2s
9:48⌁
MCP — github · create_pull_request
1.1s
9:50❯
Agent Response 41 passed. Opened PR #1184 with the migration, the writer change and the tests, and both payloads go through scrub().
Captured❮Prompts 2❯Responses 2⌁Tool calls 5⌁Skills 1⑂Subagents 1±File diffs 4

Measured impact

“We had the tools for a year. In just a couple weeks things were completely different.”

Head of Engineering50-person team using blueplane data

  • 65% decrease in AI spend
  • 50% increase in AI usage
  • Significant increase in agent autonomy

Your team’s next session is worth keeping.

Start capturing today. The library only gets more valuable from here.

Track your first session

Free to start · plans from $40 / month