SJKConsult
TMTRAILMYAIEARLY ACTIVE BUILD · PREVIEW OF THE WORKING PROTOTYPE

Your agents wrote it. Here's the flight recorder.

Every agent session, worktree, test run, review and decision lands on one timeline. See what your agents did, why they did it, and what they uncovered along the way.

LOCAL-FIRST · YOUR REPO, YOUR MACHINE
SESSION 0412 · feat/login · 2h 14mREC
TIMELINEDECISIONS 0UNCOVERED0
09:0009:3010:0010:3011:00
AGENT
GIT
TESTS
REVIEW
DECIDE
09:04
AGENT09:05refactor auth middlewareWHAT Moved token checks into one middleware. 11 files touched.WHY Session tokens were leaking into request logs.
- log.info(req.headers)+ log.info(redact(req.headers)) 11 files · +184 −122
0 AGENT RUNS0 COMMITS0 TESTS0 FINDINGS0 DECISIONSDRAG THE TIMELINE · CLICK AN EVENT
HOW IT WORKS

Three steps.

Plug it into your agent setup. Every action lands on one timeline a reviewer can scrub.

  1. 01

    Every step on a track

    Agent actions, commits, tests and reviews on one scrubbable timeline.

  2. 02

    What and why

    Each change carries the reason the agent gave, and the diff.

  3. 03

    Decisions kept

    Architecture decisions recorded next to the code they shaped.

WHERE IT IS

Early build.

  1. PROTOTYPE
  2. EARLY BUILD · NOW
  3. BETA
  4. LIVE

An early build we use on our own projects with agent work. Expect rough edges and frequent changes.

QUESTIONS

Fair questions.

No. TrailMyAI runs locally or in your cloud account.

BUILT INSIDE SJK CONSULT

Made by the people who build for clients.

We use AI agents on our own code, and wanted to review their work properly. TrailMyAI is that tool, in an early build we use ourselves.

Meet the studio
AI USE
Records and summarises. Approves nothing on its own.
DATA
Runs locally or in your cloud. Your code stays there.

Review a day of agent work, properly.

Early access opens in small groups. Get on the list for the next one.

Join the list ↑