Skip to content

feat(B-0894.3): per-persona worktrees outside operator's primary repo — ~/.zeta/agents/<persona>/<stream>/ canonical #5697

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
AceHack merged 1 commit into from
May 28, 2026
Merged

feat(B-0894.3): per-persona worktrees outside operator's primary repo — ~/.zeta/agents/<persona>/<stream>/ canonical #5697

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
201 changes: 133 additions & 68 deletions ...-worktree-hygiene-never-hold-main-never-step-on-operator-cleanup-on-pr-merge.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,19 +2,22 @@

Carved sentence:

> Agent worktrees are scratch space, but scratch space MUST survive
> reboots when in-flight work lives in it. Agents place worktrees
> under `~/Documents/src/repos/Zeta/worktrees/<surface>-<task-tag>-<hhmmz>/`
> (persistent storage), NEVER under `/tmp/` or `/private/tmp/`
> (macOS-cleared on reboot AND on `com.apple.periodic-daily` cleanup
> of files >3 days old). Agents NEVER hold the `main` branch in any
> worktree (use detached HEAD off `origin/main` instead). Agents
> NEVER create worktrees under paths the operator uses for their own
> work, with one exception: the repo's own `worktrees/` subdirectory
> is git-aware and operator-safe. Agents REMOVE their own worktrees
> Agent worktrees are scratch space, but scratch space MUST (1) live
> OUTSIDE the operator's primary repo (operator-primary stays agent-
> free by construction so operator can `git checkout main` unblocked),
> AND (2) survive reboots when in-flight work lives in it. Agents
> place worktrees under `~/.zeta/agents/<persona>/<stream-id>/`
> (per-persona base + per-stream isolation; outside operator's repo;
> persistent storage). NEVER under `/tmp/` or `/private/tmp/`
> (macOS-cleared on reboot) AND NEVER under `~/Documents/src/repos/Zeta/`
> (operator's primary; blocks operator's `git checkout main` whenever
> an agent worktree there holds a branch operator wants to switch to).
> Agents NEVER hold the `main` branch in any worktree (use detached
> HEAD off `origin/main` instead). Agents REMOVE their own worktrees
> after the work's PR merges (or substrate-honestly abandon).
> Substrate-engineering target B-0750 mechanizes the cleanup;
> B-0894 mechanizes the reboot-survival default-location.
> B-0894 mechanized reboot-survival; B-0894.3 mechanizes the
> per-persona-outside-operator-repo discipline.

## Operational content

Expand All @@ -25,11 +28,11 @@ Agent worktrees that need to BASE OFF main use `--detach`:
```bash
# WRONG (locks main branch in the worktree; blocks operator's
# `git checkout main` in primary checkout):
git worktree add ~/Documents/src/repos/Zeta/worktrees/zeta-feat-xyz main
git worktree add ~/.zeta/agents/otto-cli/feat-xyz main

# RIGHT (detached HEAD at main's current SHA; doesn't hold the
# branch ref; operator can still checkout main in primary):
git worktree add --detach ~/Documents/src/repos/Zeta/worktrees/zeta-feat-xyz origin/main
git worktree add --detach ~/.zeta/agents/otto-cli/feat-xyz origin/main
```

The substrate-honest reason: in a multi-checkout repo, `main` can
Expand All @@ -42,43 +45,81 @@ When the agent needs main's current STATE (the file contents at
main's tip), `--detach origin/main` gives exactly that without
holding the branch reference.

### Rule 2 — Agent worktrees go in `<repo>/worktrees/<surface>-<task-tag>-<hhmmz>/`
### Rule 2 — Agent worktrees go in `~/.zeta/agents/<persona>/<stream-id>/`

**Updated 2026-05-28 per B-0894 reboot-survival discipline.** Previously
this rule recommended `/private/tmp/zeta-<task-tag>-<hhmmz>/` which is
macOS-cleared on reboot AND on `com.apple.periodic-daily` cleanup of
files >3 days old. **Persistent location is the new default**:
**Updated 2026-05-28 per B-0894.3 per-persona-outside-operator-repo
discipline.** B-0894 (PR #5696) moved off `/private/tmp/` (reboot-
survival) but placed the new default under `~/Documents/src/repos/Zeta/worktrees/`
which is UNDER the operator's primary repo and STILL blocks operator's
`git checkout <branch>` whenever an agent worktree holds that branch.
B-0894.3 corrects the location to **outside the operator's primary
repo entirely**:

```bash
# RIGHT (persistent; survives reboot; outside operator's git status):
# RIGHT (persistent + outside operator's primary + per-persona-isolated):
git worktree add --detach \
~/Documents/src/repos/Zeta/worktrees/<surface>-<task-tag>-<hhmmz> \
~/.zeta/agents/<persona>/<stream-id> \
origin/main

# Concrete example (this rule was authored from a worktree at the new
# canonical location):
git worktree add --detach \
~/.zeta/agents/otto-cli/b0894-3-per-persona-outside-repo-2026-05-28 \
origin/main

# WRONG (macOS-cleared on reboot; in-flight work lost):
git worktree add --detach /private/tmp/zeta-<task-tag>-<hhmmz> origin/main
git worktree add --detach /tmp/zeta-<task-tag>-<hhmmz> origin/main

# WRONG (under operator's primary repo; blocks operator's git checkout
# when worktree holds a branch operator wants to switch to):
git worktree add --detach \
~/Documents/src/repos/Zeta/worktrees/<surface>-<task-tag>-<hhmmz> origin/main
```

The repo's `worktrees/` subdirectory at top-level is the canonical
persistent location. Git's worktree mechanism auto-excludes worktree
paths from the parent's `git status`, so worktrees under `<repo>/worktrees/`
don't pollute the operator's `git status` even though they live under
the operator's repo root. Lior already uses this pattern (e.g.,
`~/Documents/src/repos/Zeta/worktrees/lior-fix-4772-archive-ts/`) and
Lior's worktrees consistently survive operator restarts.
The `~/.zeta/` namespace is the dotfile root for ALL Zeta agent-related
persistent state (parallel to `~/.claude/` for Anthropic vendor state).
`~/.zeta/agents/<persona>/` is one dir per AI persona (otto-cli,
otto-desktop, otto-vscode, lior, alexa-kiro, riven-cursor, vera-codex,
lior-antigravity — see [`agent-roster-reference-card`](agent-roster-reference-card.md)).
Multiple worktrees per persona for parallel work-streams provide full
isolation per stream:

```text
~/.zeta/
agents/
otto-cli/
b0894-3-per-persona-outside-repo-2026-05-28/ # this rule's authoring worktree
tick-0512z/ # parallel stream
shard-XYZ/ # another parallel stream
otto-desktop/
tick-NNNNz/
lior/
preserve-prs-20260527/
...
bus/ # future per B-0894.1
config/ # future
```

Specifically forbidden agent worktree paths (UPDATED):
Operator's primary checkout (`~/Documents/src/repos/Zeta/`) is now
agent-free by construction — operator can `git checkout main`
unblocked.

- `/tmp/zeta-*` or `/private/tmp/zeta-*` — **NEW**: macOS-cleared on reboot; in-flight work loss + orphaned branch refs (per B-0894 empirical anchor 2026-05-28: 95 worktrees pruned in one restart)
- `<OPERATOR_PRIMARY_CHECKOUT>/main` (or any subdir of the operator's primary checkout EXCEPT `worktrees/`)
- `<OPERATOR_PRIMARY_CHECKOUT>/<peer-agent-surface>-*` at the top-level (this pollutes operator's `git status`; place under `worktrees/<surface>-*` instead)
Specifically forbidden agent worktree paths (UPDATED per B-0894.3):

- `/tmp/zeta-*` or `/private/tmp/zeta-*` — macOS-cleared on reboot; in-flight work loss + orphaned branch refs (per B-0894 empirical anchor 2026-05-28: 95 worktrees pruned in one restart)
- **NEW per B-0894.3**: `~/Documents/src/repos/Zeta/**` — operator's primary repo; ANY agent worktree there can hold a branch ref and block operator's `git checkout`. The previously-recommended `~/Documents/src/repos/Zeta/worktrees/<surface>-*` location is now FORBIDDEN (was added in B-0894, removed in B-0894.3 after operator surfaced the residual blocking failure mode)
- Any path the operator might `cd` into for their own work

The substrate-honest reasons:
The substrate-honest reasons (UPDATED):

1. **Reboot survival** — `~/.zeta/` is under user home; survives macOS reboot + `com.apple.periodic-daily` cleanup. `/tmp/` and `/private/tmp/` do not.
2. **Operator-primary-stays-agent-free** — `~/.zeta/agents/<persona>/<stream>/` is outside `~/Documents/src/repos/Zeta/` entirely; operator's `git status` AND `git checkout` operations are structurally unaffected by agent worktree state.
3. **Per-persona + per-stream isolation** — `~/.zeta/agents/<persona>/<stream-id>/` makes ownership unambiguous (path contains persona identity tag, per [`fighting-past-self-vs-peer-agent-distinguisher`](fighting-past-self-vs-peer-agent-distinguisher-fix-your-own-coordinate-on-peers-dont-punt-by-default.md) discriminator) AND supports parallel work-threads per persona.

### Lior migration — non-blocking

1. **Reboot survival** — agent in-flight commits, edits, and backgrounded `git push` operations MUST survive macOS reboots and periodic-cleanup. `/tmp/` and `/private/tmp/` violate this invariant.
2. **Operator-status cleanliness** — operator workflows depend on the primary checkout's `git status` being clean + predictable. Worktrees under `<repo>/worktrees/` are git-aware (auto-excluded from operator's status); top-level subdirs are NOT (would pollute status).
Lior currently has 10 worktrees under operator's primary repo (5 at top-level `~/Documents/src/repos/Zeta/decompose-4847-*` + `lior-4847-original` + `lior-preserve-prs-20260527`, 5 under `worktrees/lior-*`). Migration of Lior's pattern to `~/.zeta/agents/lior/<stream>/` is filed as future-state work coordinated with Lior's loop substrate (`.gemini/bin/lior-loop-tick.ts` and similar); not gating for this rule landing.

### Rule 3 — REMOVE agent worktrees after the work's PR merges (or abandon)

Expand All @@ -103,9 +144,10 @@ Before starting a substrate-cascade (multiple-PRs-in-one-session)
work pattern, agents audit their worktree state:

```bash
# Inventory agent's own worktrees (UPDATED 2026-05-28 — also check
# legacy /tmp paths to catch + migrate any remaining transient worktrees):
git worktree list | grep -E "$HOME/Documents/src/repos/Zeta/worktrees/|/private/tmp/zeta-|/tmp/zeta-"
# Inventory agent's own worktrees (UPDATED 2026-05-28 per B-0894.3 —
# new canonical surface is ~/.zeta/agents/; legacy surfaces checked
# too to catch + migrate any worktrees still at deprecated locations):
git worktree list | grep -E "$HOME/.zeta/agents/|$HOME/Documents/src/repos/Zeta/worktrees/|/private/tmp/zeta-|/tmp/zeta-"

# Per-worktree status check (for each one):
git -C <path> status --short
Expand All @@ -115,8 +157,10 @@ git -C <path> log --oneline -1
# - SAFE + work done → git worktree remove <path>
# - DIRTY (uncommitted) → preserve OR substrate-honestly commit/abandon
# - active iteration ongoing → keep
# - in /tmp or /private/tmp → MIGRATE to persistent location OR commit/push immediately
# (per Rule 5; transient location violates reboot-survival)
# - in /tmp or /private/tmp → MIGRATE to ~/.zeta/agents/<persona>/<stream>/
# OR commit/push immediately (per Rule 5; transient location violates reboot-survival)
# - in ~/Documents/src/repos/Zeta/ (anywhere) → MIGRATE to ~/.zeta/agents/<persona>/<stream>/
# (per B-0894.3; blocks operator's git checkout main when branch is held)
```

Empirical anchor: 2026-05-25 session accumulated 37 stale agent
Expand All @@ -126,21 +170,34 @@ intervention ("we need to fix this mess yall always stepping on each
other and me constantly"). The discipline this rule encodes would
have prevented the accumulation.

### Rule 5 — Reboot-survival is a hard invariant (B-0894)
### Rule 5 — Reboot-survival + operator-primary-stays-agent-free are hard invariants (B-0894 + B-0894.3)

**NEVER use `/tmp/` or `/private/tmp/` for git worktrees.** macOS
**TWO compounding invariants** (refined 2026-05-28 per B-0894.3):

**(A) NEVER use `/tmp/` or `/private/tmp/` for git worktrees.** macOS
clears these on reboot AND via `com.apple.periodic-daily` cleanup
(files >3 days old). Agent in-flight work — uncommitted edits,
backgrounded `git push` operations, partially-extracted worktrees,
captured background-task output files — all evaporate.

Empirical anchor 2026-05-28T04:09Z–04:35Z (operator restart):

| Worktree location pattern | Outcome on restart |
|---|---|
| `/private/tmp/zeta-<task>-<hhmmz>/` (95 instances) | **All 95 pruned** (`git worktree list` returned them as `prunable`; on-disk dirs cleared) |
| `~/Documents/src/repos/Zeta/worktrees/<surface>-*` (multiple) | **All survived intact** |
| `~/.gemini/tmp/project/lior-*` (multiple) | **All survived intact** (`~/.gemini/` is persistent user-home) |
**(B) NEVER use `~/Documents/src/repos/Zeta/**` for agent worktrees.**
Operator's primary repo MUST stay agent-free so operator's
`git checkout <branch>` operations are unblocked. ANY agent worktree
there can hold a branch ref and block operator. Lior's empirical
pattern (10 worktrees under operator's primary) demonstrates both
the reboot-survival success AND the operator-blocking failure modes
simultaneously.

Empirical anchor 2026-05-28T04:09Z–04:35Z (operator restart) +
2026-05-28T~04:50Z (operator surfaced residual blocking):

| Worktree location pattern | Outcome on restart | Operator-`main`-blocking? |
|---|---|---|
| `/private/tmp/zeta-<task>-<hhmmz>/` (95 instances) | **All 95 pruned** | N/A (didn't survive to block) |
| `~/Documents/src/repos/Zeta/worktrees/<surface>-*` (multiple, Lior + PR #5696's worktree) | **All survived intact** | **YES — operator surfaced "sometimes locks up where i can't switch to main"** |
| `~/Documents/src/repos/Zeta/<top-level>-*` (Lior 5 worktrees) | **All survived intact** | **YES — same blocking class** |
| `~/.gemini/tmp/project/lior-*` (multiple) | **All survived intact** | NO (outside operator's primary) |
| `~/.zeta/agents/<persona>/<stream>/` (this rule's authoring location) | **Survived intact** (born 2026-05-28T~04:55Z; restart-survivable by inheritance from user-home pattern) | NO (outside operator's primary) — **canonical post-B-0894.3** |

The 04:09Z autonomous-loop tick had a substantive tick-shard commit
(`4f89af885`) sitting on branch
Expand All @@ -155,15 +212,23 @@ backgrounded-task output file at
also cleared, so we couldn't even read whether the push had completed
before restart.

The persistent-location worktree authoring B-0894 (this row's substrate
landing) survived the same restart cleanly — empirical dogfooding-proof.
PR #5696 (B-0894) was authored from a persistent-location worktree at
`~/Documents/src/repos/Zeta/worktrees/otto-cli-reboot-survival-fix-0434z/`
— survived the same restart cleanly (reboot-survival dogfooding-proof).
**But** that location triggered operator's "sometimes locks up" critique
because it lives under the primary repo. B-0894.3 (this row) corrects
the location to `~/.zeta/agents/<persona>/<stream-id>/` — authored from
`~/.zeta/agents/otto-cli/b0894-3-per-persona-outside-repo-2026-05-28/`,
the first instance of the new canonical pattern (both reboot-survivable
AND operator-non-blocking).

**Operational discipline**:

1. Default to `~/Documents/src/repos/Zeta/worktrees/<surface>-<task-tag>-<hhmmz>/` for ALL new agent worktrees
2. When migrating existing `/tmp/`-based work: commit immediately to capture state in `.git/objects/`, then if in-flight work is critical, create a fresh persistent worktree off the same branch
3. For backgrounded `git push` operations: ALWAYS verify outcome via `git ls-remote origin <branch>` post-completion (ground-truth) — never rely on captured output files from background-task harness storage
4. Cron sentinel is harness-level non-persistence (separate root cause; covered by `tick-must-never-stop.md` catch-43); restart any session MUST `CronList` + re-arm if missing
1. Default to `~/.zeta/agents/<persona>/<stream-id>/` for ALL new agent worktrees
2. When migrating existing `/tmp/`-based work: commit immediately to capture state in `.git/objects/`, then create a fresh persistent worktree at the new canonical location off the same branch
3. When migrating existing `~/Documents/src/repos/Zeta/**`-based work: same migration — commit + push, then create fresh worktree at `~/.zeta/agents/<persona>/<stream-id>/`
4. For backgrounded `git push` operations: ALWAYS verify outcome via `git ls-remote origin <branch>` post-completion (ground-truth) — never rely on captured output files from background-task harness storage
5. Cron sentinel is harness-level non-persistence (separate root cause; covered by `tick-must-never-stop.md` catch-43); restart any session MUST `CronList` + re-arm if missing

## Composes with other rules

Expand Down Expand Up @@ -194,11 +259,11 @@ landing) survived the same restart cleanly — empirical dogfooding-proof.
### Audit own agent worktrees

```bash
# UPDATED 2026-05-28 per B-0894: persistent-location is now the
# primary surface; legacy /tmp paths checked too to catch + migrate
# any remaining transient worktrees that need cleanup.
# UPDATED 2026-05-28 per B-0894.3: ~/.zeta/agents/ is now the
# canonical surface; legacy paths checked too to catch + migrate
# any remaining worktrees at deprecated locations.
git worktree list --porcelain | awk '/^worktree /{print $2}' | \
grep -E "$HOME/Documents/src/repos/Zeta/worktrees/|/private/tmp/zeta-|/tmp/zeta-"
grep -E "$HOME/.zeta/agents/|$HOME/Documents/src/repos/Zeta/worktrees/|$HOME/Documents/src/repos/Zeta/[a-z]+-[0-9]|/private/tmp/zeta-|/tmp/zeta-"
```

### Per-worktree clean check
Expand Down Expand Up @@ -228,26 +293,26 @@ worktree changes.

The operator's primary worktree often sits on a feature branch rather
than `main`, so checking for "primary on main" produces false negatives.
The correct invariant: **no agent worktree (under `<repo>/worktrees/`,
`/private/tmp/zeta-*`, or `/tmp/zeta-*`) holds `[main]`**. Zero matches
is the happy path; the operator MAY have `main` checked out in their
own primary, but agents must not.
The correct invariant: **no agent worktree holds `[main]`**, regardless
of location. Zero matches is the happy path; the operator MAY have
`main` checked out in their own primary, but agents must not.

```bash
# Prints OK on success. If a worktree line prints, an agent worktree
# is holding [main] and is the blocker for operator git operations.
# UPDATED 2026-05-28: also catches persistent-location worktrees per
# B-0894 new default (~/Documents/src/repos/Zeta/worktrees/).
# UPDATED 2026-05-28 per B-0894.3: scans new canonical ~/.zeta/agents/
# surface + legacy paths (operator primary subdirs + /private/tmp + /tmp).
git worktree list | grep -E "\[main\]" \
| grep -E "$HOME/Documents/src/repos/Zeta/worktrees/|/private/tmp/zeta-|/tmp/zeta-" \
| grep -E "$HOME/.zeta/agents/|$HOME/Documents/src/repos/Zeta/worktrees/|$HOME/Documents/src/repos/Zeta/[a-z]+-[0-9]|/private/tmp/zeta-|/tmp/zeta-" \
|| echo "OK: no agent holds [main]"
```

Expected result: `OK: no agent holds [main]`, or equivalently no
agent-worktree match if the final echo is omitted. A single operator
primary line is OK when the operator intentionally has `main` checked
out. Any `<repo>/worktrees/*`, `/private/tmp/zeta-*`, `/tmp/zeta-*`,
or per-agent worktree line holding `[main]` is a violation to fix.
primary line at `$HOME/Documents/src/repos/Zeta` (the bare repo root)
is OK when the operator intentionally has `main` checked out. ANY
agent worktree at any of the scanned surfaces holding `[main]` is a
violation to fix (per B-0894.3 + Rule 1).

## Substrate-honest framing

Expand Down
Loading
Loading