Log every step, including Thinking and Tool calls

Agent session logs are stored and searchable. You can even download the raw logs and push your local Claude Code sessions to the platform.

01

Status and outcomes at a glance

One live view shows every session across your org: what's running now, what shipped, and what failed, with durations and outcomes by agent and repo.

Sessionslive
Flagged 1 risk on api-repo #1841
Dropped column without backfill
$0.21
12.0K tok
just now1m30s
Posted daily standup to your-org #engineering
4 merged, 2 awaiting review
$0.05
3.3K tok
just now22s
Opened web-repo #314
Implement CSV export
$0.42
31.2K tok
22s ago4m02s
Approved api-repo #1839
Add rate limiting to /v1
$0.18
9.1K tok
1m ago48s
Reviewed web-repo #1842
Fix connection leak on error path
$0.31
18.4K tok
3m ago1m12s

Every session across the org, live

02

Know what every session cost

Each session reports its exact cost, broken down by tokens, CPU, and memory. Chart spend over time and see it by agent and by repo, so you always know where the money goes.

Spend by agentlast 7 days
code-reviewer$14.20
issue-to-pr$9.84
dep-updater$4.31
standup-bot$1.62
session #1842 · $0.31tokens $0.27 · CPU $0.03 · memory $0.01total $29.97

Spend by agent, tokens, CPU, and memory

03

Watch a session while it works

Follow any session in real time: steps stream to the CLI or dashboard as they happen, and a live summary says what the agent is doing right now. The same WebSocket feeds your own tools through the API.

Terminalstreaming
$ agent session get session_7Hq2mX4p --watch
running · issue-to-pr · web-repo
Read web/exports/csv.ts
Edit web/exports/csv.ts · +84 −0
Bash pnpm test exports · 12 passed
live: opening a PR for the CSV export
completed · $0.42 · 4m02s · PR #314

Steps stream to the CLI as they happen

04

See every step, from Thinking to Diff

When a session needs a closer look, open it: the agent's thinking, the tools it called, the tests it ran, and the diff it shipped, with per-step token counts. Every raw transcript is kept as a durable record, so Ellipsis is the system of record for your agent work.

Steps (15)
15Commit pushed28 tok
14thinkingDouble-check the diff before pushing
13Bashgit commit -m 'Release connection on error path'20 tok
12thinkingWrite a clear commit message
11assistantGreen. Committing the fix.14 tok
10Bashpytest -k pool · 142 passed24 tok
9thinkingRe-run the pool suite to confirm the fix
Budget$0.40 / $1.00
Tokens30.8K

The full step timeline, kept permanently

05

Search everything your agents have ever done

Transcripts, recaps, and PR references are all indexed, with semantic search alongside full text, so did-anyone-look-into-this is a query, not archaeology.

$ agent session search "webhook retries"

Investigated webhook retry storms after the queue migration

session_4Hn2… · tomas-vidal · api-repo · 3 days agostepsrecap

Added idempotency keys to the trade-webhook consumer

session_8Rw9… · ellipsis[bot] · api-repo · last weeksimilarpr
full text + semantic similarity, one query

Semantic search over every transcript

The complete session record

Everything the platform records and keeps, so agent work has the same paper trail as the rest of your infrastructure.

Every step

Assistant output, each tool call with its input and result, and per-step token counts, recorded live.

Exact cost

Total spend broken into model tokens, sandbox CPU, sandbox memory, and the fee, with per-turn timing.

The config snapshot

The exact config the session executed. Later edits never change what a past session shows.

The trigger

The person or event that started every session, so attribution is on the record.

Exit statuses

A budget cutoff reads differently from a crashed tool call: budget_hit, tool_call_failed, and friends.

Downloadable logs

The complete ordered session history, one JSONL file from the CLI or API.

Live streaming

Sessions stream to the dashboard, the CLI with --watch, and your own tools over WebSocket.

Steering

Message a running session and it redirects without restarting; stop it from the dashboard or CLI.

Search

Semantic search alongside full text across transcripts, recaps, and PR references.

Retention you control

Transcripts keep indefinitely by default; admins can set retention from 7 to 365 days.

API access

Everything the dashboard shows is readable over the REST API: sessions, steps, costs, and logs.

Frequently asked questions

Explore the platform