traqr.dev·working document·sheet 03 — cli
One repo, several Claude sessions, no collisions.
The CLI is how a project gets its shape. traqr init interviews you and writes .traqr/config.json; traqr rendergenerates everything downstream — the CLAUDE.md, the skills, the worktree slots, the Guardian config. The slots are the part I lean on hardest: each one is a git worktree with its own branch, port, and Claude session, so parallel work doesn’t step on itself.
§ 00·the slot system
A worktree per workstream.
The repo this site describes carries 18 slots — five each for features, fixes, and infra, two ambient on a Mac Mini, one for marketing. On a normal night two to four are active. Shell aliases jump between them; a pre-push hook keeps everyone — including me — off main.
fig. 1 — a slot session, and the hook doing its job. Counted from .worktrees/ on 2026-06-09.
§ 01·shipping
/ship, then walk away.
One command moves work from a slot to main. Everything that used to be a checklist — rebase, build, PR, notify — runs as a documented procedure, and Guardian owns the second half.
commit + rebase
Commits what's staged, rebases onto main, and runs an impact analysis on anything that overlaps — risky overlap stops the ship.
build
Builds the shared packages, then only the apps the diff actually touches. A failing build stops here, not in CI.
push + open the PR
Pushes the slot branch and opens the pull request with the label Guardian watches for.
notify
Posts the PR to Slack and captures what the session learned into memory.
hand off to Guardian
Rebase, merge, sync the other slots. By the time I look, it's on main.
§ 02·skills
The index.
Skills are slash commands — markdown procedures the AI follows, generated from your config. This repo carries 57 of them, counted 2026-06-09. These are the ones I use most; each is a file you can read, edit, or delete.
/shipcommit, build, push, open the PR, notify Slack/resyncpre-flight rebase with impact analysis/syncsync every worktree slot after a merge/simplifyreview changed code for reuse and efficiency/nextphasepick up the next phase of a Linear project/einsteindeep research with parallel subagent waves/bethesdaautonomous exploration — agents find their own work/debatestructured devil's advocacy on a decision/deepreflectmeta-analysis of how the agents are performing/coschief-of-staff awareness across all slots/daily-briefmorning intelligence from every data source/callprocess a meeting transcript into structured notes/checkpointmid-work save to memory and the vault/deciderecord a decision with its full context/traqr-initinitialize TraqrOS on any project/traqr-updateupgrade in place, preserving your sections/traqr-system-slackwire Slack routing and Block Kit/traqr-system-guardianconfigure the automated PR lifecycle/traqr-system-cronsstandardized cron auth and escalation…and ~30 more. Each one is a markdown file you can read, edit, or delete.
§ 03·install
Four commands.
init once, render after config changes, status and verify when something feels off. The memory server installs standalone if you only want that piece.
traqr initinterview your project, write .traqr/config.jsontraqr rendergenerate the harness from the configtraqr statuswhat's configured, what's driftedtraqr verifycheck the generated pieces still match the config