release: v1.4.0 — typst PDF engine + host SSH startup check

Finalize the Unreleased batch as v1.4.0 (minor — significant base
additions). Base rebuilds (Dockerfile.base for typst/xz-utils;
.bash_aliases for the SSH check). pi auto-resolves latest (0.80.3 ->
0.80.6); mempalace stays 3.5.0 (current).

.env.example

@@ -9,6 +9,19 @@ WORKSPACE_PATH=~/projects
 # Path to SSH keys on host
 SSH_KEY_PATH=~/.ssh
 
+# ── MemPalace memory (local by default) ───────────────────────────
+# By default the mempalace.ts extension spawns a LOCAL mempalace-mcp stdio
+# server (palace at ~/.mempalace). Uncomment the devbox-palace volume in
+# docker-compose.yml to persist it across container recreation.
+#
+# To instead share ONE MemPalace across containers/harnesses (pi + opencode
+# + native), set the URL below. When set, the extension connects over HTTP
+# and NO local mempalace-mcp is spawned; the devbox-palace volume is then
+# irrelevant. MEMPALACE_REMOTE_TOKEN, if set, is sent as a bearer token.
+#   Serve it with:  mempalace-mcp --transport http --host 0.0.0.0 --port 8765
+# MEMPALACE_REMOTE_URL=http://mempalace.lan:8765/mcp
+# MEMPALACE_REMOTE_TOKEN=
+
 # ── LAN access from the container (host-OS-agnostic) ─────────────────
 # On VM-backed hosts (macOS OrbStack / Docker Desktop) the container can't
 # reach the host's directly-attached LAN peers by default. The entrypoint

.gitea/workflows/docker-publish.yml

@@ -38,6 +38,16 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: false
 
+# Gitea Actions' default step shell is `sh -e {0}` (dash), which rejects
+# bash-only syntax like `set -o pipefail`, `[[ ]]`, and arrays. Setting the
+# default to bash workflow-wide eliminates the whole class of "forgot
+# `shell: bash` on this step" bugs (hit twice: ed49b8d resolve-versions,
+# b7197e8/b33e9dc promote-base-latest). All existing dash steps use only
+# POSIX syntax, so bash (a superset) runs them unchanged.
+defaults:
+  run:
+    shell: bash
+
 env:
   BUILDKIT_PROGRESS: plain
   IMAGE: ${{ vars.DOCKERHUB_USERNAME }}/pi-devbox
@@ -565,16 +575,19 @@ jobs:
     needs:
       - base-decide
       - build-variant
-    # Skip on cache-hit base builds: when need_build=false, base-latest
-    # already points at the same digest as base-<hash>, so the retag is
-    # a tautology and any transient failure of it is purely cosmetic.
-    # Manual workflow_dispatch with promote_latest=true overrides this
-    # gate as an escape hatch (e.g., if base-latest got hand-deleted).
+    # Run on every tag release (and on promote_latest=true dispatches).
+    # The job-level gate deliberately does NOT key off need_build anymore:
+    # the actual no-op optimization moved INTO the step as a digest compare
+    # (see below). Keying the gate on need_build was wrong because a prior
+    # dry-run dispatch (promote_latest=false) can pre-build+push base-<hash>,
+    # making need_build=false on the subsequent tag run even though
+    # base-latest is still stale — the old gate then skipped promotion and
+    # left base-latest pointing at the PREVIOUS base. (Observed 2026-06-27,
+    # v1.2.3: dry-run-first release left base-latest one base behind.)
     if: |
       always() &&
       needs.build-variant.result == 'success' &&
-      (inputs.promote_latest == 'true' ||
-       (github.ref_type == 'tag' && needs.base-decide.outputs.need_build == 'true'))
+      (inputs.promote_latest == 'true' || github.ref_type == 'tag')
     runs-on: ubuntu-latest
     container:
       image: catthehacker/ubuntu:act-latest
@@ -596,11 +609,38 @@ jobs:
           crane auth login docker.io \
             -u ${{ vars.DOCKERHUB_USERNAME }} \
             -p "${{ secrets.DOCKERHUB_TOKEN }}"
-      - name: Re-tag base-<hash> as base-latest
+      - name: Re-tag base-<hash> as base-latest (only if stale)
+        # shell: bash is REQUIRED — Gitea Actions' default step shell is
+        # `sh -e {0}` (dash), which rejects `set -o pipefail` with
+        # "Illegal option -o pipefail" and aborts the step before the
+        # crane digest-compare runs, leaving base-latest un-promoted.
+        # Same footgun as ed49b8d (resolve-versions). Regression shipped
+        # in b7197e8, caught on the v1.2.4 release (run 418).
+        shell: bash
+        env:
+          BASE_HASH_REF: ${{ env.IMAGE }}:${{ needs.base-decide.outputs.base_tag }}
+          BASE_LATEST_REF: ${{ env.IMAGE }}:base-latest
         run: |
-          crane copy \
-            ${{ env.IMAGE }}:${{ needs.base-decide.outputs.base_tag }} \
-            ${{ env.IMAGE }}:base-latest
+          set -euo pipefail
+          # Correctness invariant: after a release, base-latest must resolve to
+          # the SAME digest as the base-<hash> the just-built variants were
+          # FROM. Compare digests rather than trusting need_build — a prior
+          # dry-run dispatch can pre-build base-<hash>, so need_build=false on
+          # the tag run does NOT imply base-latest is already current. When the
+          # digests already match (genuine cache-hit release) this is a no-op,
+          # so we skip the crane copy entirely — preserving the original
+          # "don't do a tautological retag" intent and avoiding any cosmetic
+          # transient-failure exposure on releases that change nothing.
+          want=$(crane digest "${BASE_HASH_REF}")
+          have=$(crane digest "${BASE_LATEST_REF}" 2>/dev/null || echo "")
+          echo "base-<hash> digest: ${want}"
+          echo "base-latest digest: ${have:-<absent>}"
+          if [ "${want}" = "${have}" ]; then
+            echo "base-latest already current; nothing to promote."
+          else
+            echo "Promoting base-latest -> ${BASE_HASH_REF}"
+            crane copy "${BASE_HASH_REF}" "${BASE_LATEST_REF}"
+          fi
 
   # ── Phase 6: update Hub description (only on real release runs) ────
   update-description:

.gitea/workflows/lint.yml

@@ -0,0 +1,69 @@
+name: Lint workflows
+
+# Durable guard against CI-workflow bugs — most importantly the recurring
+# "bash-only syntax under the default `sh`/dash shell" footgun that broke
+# resolve-versions (ed49b8d) and promote-base-latest (b7197e8 → run 418).
+# actionlint runs shellcheck against each `run:` step using its *effective*
+# shell, so `set -o pipefail` under dash is flagged as SC3040 before any
+# expensive build runs. This is cheap (~10s) and independent of the build
+# pipeline, so it fires on every push/PR — not just on release tags, which
+# is where the build workflow (docker-publish.yml) is otherwise only
+# triggered.
+on:
+  push:
+  pull_request:
+  workflow_dispatch:
+
+concurrency:
+  group: lint-${{ github.ref }}
+  cancel-in-progress: true
+
+defaults:
+  run:
+    shell: bash
+
+jobs:
+  actionlint:
+    runs-on: ubuntu-latest
+    container:
+      image: catthehacker/ubuntu:act-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install shellcheck
+        run: |
+          apt-get update
+          apt-get install -y --no-install-recommends shellcheck python3-yaml
+
+      - name: Gitea shell guard (catches the actionlint blind spot)
+        # actionlint models GitHub Actions, where the default run shell is
+        # bash, so it does NOT flag bash syntax in a step that merely OMITS
+        # `shell:` — which is exactly how ed49b8d and b7197e8 manifested on
+        # Gitea (default sh/dash). This guard enforces that every run: step
+        # resolves to bash under Gitea's real defaults. Run it BEFORE
+        # actionlint so the more precise diagnostic surfaces first.
+        run: bash scripts/check-workflow-shell.sh .gitea/workflows
+
+      - name: Install actionlint (pinned)
+        env:
+          ACTIONLINT_VERSION: 1.7.7
+        run: |
+          curl -fsSL \
+            "https://github.com/rhysd/actionlint/releases/download/v${ACTIONLINT_VERSION}/actionlint_${ACTIONLINT_VERSION}_linux_amd64.tar.gz" \
+            | tar -xz -C /usr/local/bin actionlint
+          actionlint --version
+
+      - name: Run actionlint
+        # SHELLCHECK_OPTS excludes pure-style codes (quoting/style opinions)
+        # so the guard stays focused on correctness bugs — crucially the
+        # SC3xxx "not POSIX / wrong shell" family that catches the pipefail
+        # footgun. Do NOT exclude SC3040 (set -o pipefail under sh) or any
+        # other SC3xxx code.
+        env:
+          SHELLCHECK_OPTS: "-e SC2086 -e SC2016 -e SC2129 -e SC2001 -e SC2312"
+        # Pass explicit paths: actionlint's no-arg mode auto-detects a
+        # project by looking for `.github/workflows`, which doesn't exist in
+        # this `.gitea/workflows` repo and hard-fails with exit 3
+        # ("no project was found"). Globbing the workflow files is the
+        # supported way to lint a non-GitHub layout.
+        run: actionlint -color .gitea/workflows/*.yml

AGENTS.md

@@ -153,10 +153,14 @@ deprecated artifacts (to be removed in opencode-devbox v2.0.0).
 
 ## What we DON'T install (and why)
 
-- **No texlive** (~600 MB–1 GB). Users who need PDF export from pandoc
-  or pi-studio can install on demand: `sudo apt-get install texlive-xetex
-  texlive-latex-recommended`. The planned `:latest-studio-tex` variant
-  will bake this in.
+- **No texlive** (~600 MB–1 GB). PDF export from pandoc / pi-studio works
+  out of the box via **`typst`** (~30 MB static binary), which the base ships
+  as the pandoc PDF engine (`pandoc --pdf-engine=typst`) — small enough to live
+  in base rather than a dedicated `:latest-studio-tex` variant. We don't bake in
+  a full TeX Live: it's heavy and typst covers the common Markdown→PDF case.
+  Users needing LaTeX-exact output can install the higher-fidelity fallback on
+  demand: `sudo apt-get install texlive-xetex texlive-latex-recommended` (then
+  `pandoc --pdf-engine=xelatex`).
 - **pi-studio** ships in the `:latest-studio` variant (since v1.1.0),
   vendored to `/opt/pi-studio` and registered at container start via
   `pi install /opt/pi-studio` (see Dockerfile.variant `INSTALL_STUDIO`).

CHANGELOG.md

@@ -11,6 +11,241 @@ Pre-v1.0.0 tags followed the pi npm version (`v{pi_version}[letter]`).
 
 ---
 
+## v1.4.0 — 2026-07-11
+
+Minor release. Headline: **PDF export works out of the box** — the base now
+ships **`typst`** as the pandoc PDF engine (`pandoc --pdf-engine=typst`), so
+`studio_export_pdf` / `pandoc -o out.pdf` no longer fail with "xelatex not
+found". Also adds a **host SSH reachability check at shell startup**. Both are
+base-affecting (`Dockerfile.base` apt+RUN for typst/xz-utils; `.bash_aliases`
+for the SSH check is COPYd into the base), so the base rebuilds and both land in
+`base-<hash>`. pi auto-resolves `latest` at build time (0.80.3 → 0.80.6);
+mempalace stays pinned at 3.5.0 (current PyPI latest).
+
+### Added
+
+- **Host SSH reachability check at shell startup.** `~/.bash_aliases` (baked
+  into the image) now runs a one-time SSH probe on the first bash session of
+  each container. If the Mac host is not reachable (Remote Login disabled or
+  the `devbox_jump` key not yet authorized) it prints a clear warning with the
+  exact two steps to fix it, including the container's public key inline.
+  Subsequent shells in the same container skip the check (flag in `/tmp`,
+  cleared on recreate). Silent when SSH is working. Complements the existing
+  key-generation message in `setup-lan-access.sh` which only fires once at key
+  creation time and can easily be missed. Commit `4563b4d`.
+
+- **`typst` — lightweight PDF engine for pandoc (Markdown→PDF).** `pandoc` has
+  shipped in the base since v1.0.0 but as a front-end only — with no PDF
+  back-end installed, `studio_export_pdf` / `pandoc -o out.pdf` failed with
+  "xelatex not found". The base now installs `typst`, a single ~30 MB static
+  Rust binary (no LaTeX), used via `pandoc --pdf-engine=typst`. Chosen over a
+  ~600 MB TeX Live install; a fuller TeX Live remains the higher-fidelity
+  fallback for anyone needing LaTeX-exact output (install on demand). Also adds
+  `xz-utils` to the apt layer (typst ships a `.tar.xz` asset that `tar` needs
+  `xz` to extract). Installed with the standard `latest` GitHub-release idiom;
+  pin with `--build-arg TYPST_VERSION=vX.Y.Z`. This lands in `base-<hash>`
+  (Dockerfile.base changed). Supersedes the previously-planned
+  `:latest-studio-tex` variant — typst is small enough to ship in BASE, so no
+  separate TeX variant is needed. See `pi-devbox-roadmap`.
+
+---
+
+## v1.3.0 — 2026-07-02
+
+Minor release. Headline: **shared/external MemPalace** — the `mempalace.ts`
+bridge can now point at one MemPalace HTTP server (`MEMPALACE_REMOTE_URL`,
+optional `MEMPALACE_REMOTE_TOKEN`) shared across containers/harnesses instead of
+a per-container local palace; ships `docker-compose.mempalace.yml` for the
+server. Also ships the **`nano` + `micro`** non-modal editors and a **CI
+workflow-lint layer** (Gitea-accurate sh-vs-bash guard + actionlint/shellcheck),
+with the `docker-publish.yml` bash-defaults and `promote-base-latest` shell
+fixes. pi stays `0.80.3`; the base image rebuilds (the mempalace-toolkit ref
+advanced and `Dockerfile.base` gained nano/micro), so the new bridge and editors
+land in `base-<hash>`.
+
+### Added
+
+- **Share one MemPalace across containers via `MEMPALACE_REMOTE_URL`.** The
+  `mempalace.ts` bridge (from `mempalace-toolkit`) can now connect to a shared
+  MemPalace over HTTP instead of spawning a per-container local server: set
+  `MEMPALACE_REMOTE_URL=http://<host>:8765/mcp` (optionally
+  `MEMPALACE_REMOTE_TOKEN`) in `.env` and no local `mempalace-mcp` is spawned.
+  A new `docker-compose.mempalace.yml` stands up such a shared server
+  (`mempalace-mcp --transport http`). Leaving the URL unset keeps the default
+  local-per-container palace. See `.env.example`. (The HTTP transport is
+  unauthenticated — keep it on a trusted network or behind a reverse proxy.)
+
+- **Two non-modal terminal editors alongside `nvim`: `nano` and `micro`.**
+  The image previously shipped only `nvim` (with `EDITOR=nvim`), a modal
+  vi-style editor. Not everyone is comfortable with vi keybindings, so both
+  a classic and a modern non-modal option now ship:
+  - **`nano`** (apt) — ~2.8 MB installed. Its dependencies (`libc6`,
+    `libncursesw6`, `libtinfo6`) are already present via `nvim`/`less`/`htop`/
+    `tmux`, so it pulls in **no extra packages**. On-screen shortcut hints
+    (`^O` write, `^X` exit) make it the lowest-friction fallback.
+  - **`micro`** — ~12 MB, a single static Go binary installed from GitHub
+    releases (same pattern as `bat`/`eza`/`zoxide`). Desktop-style keybindings
+    (`Ctrl+S` save, `Ctrl+Q` quit, `Ctrl+C/V/X`, `Ctrl+Z` undo), mouse
+    support, and syntax highlighting out of the box. Pin with
+    `--build-arg MICRO_VERSION=vX.Y.Z`; defaults to `latest`.
+
+  Combined footprint is ~15 MB (<0.5% of the ~3.2 GB image). **`EDITOR`
+  stays `nvim`** — the new editors are opt-in via `export EDITOR=micro`
+  (or `nano`) and/or `git config --global core.editor micro`.
+
+  Note: micro's upstream repo moved `zyedidia/micro` → `micro-editor/micro`;
+  the Dockerfile uses the canonical URL because the old org's
+  `/releases/latest` redirect lands on another `/latest` URL (the org
+  rename), which would defeat the tag-parsing `latest`-resolution idiom.
+  These are base-image additions, so they only land once the `base-<hash>`
+  rebuilds (this file changed, so the next build picks them up).
+
+### Added (CI)
+
+- **Workflow lint (`.gitea/workflows/lint.yml`) running on every push and PR.**
+  Two complementary checks, so CI-workflow bugs are caught before an expensive
+  build runs:
+  - **`scripts/check-workflow-shell.sh`** — a Gitea-accurate guard that fails
+    if any `run:` step doesn't resolve to `bash` under Gitea's real defaults.
+    This catches the exact recurrence class (omit `shell:`, use bash syntax),
+    which **actionlint alone does not** — actionlint models GitHub Actions
+    (default shell = bash) and so assumes a shell-less step is bash, whereas
+    Gitea's default is `sh`/dash.
+  - **`actionlint` + `shellcheck`** — catches explicit `shell: sh` + bash
+    syntax (SC3040 etc.), expression errors, and general workflow mistakes.
+    Style-only shellcheck codes are excluded; the SC3xxx "wrong shell" family
+    is kept.
+
+### Changed (CI)
+
+- **Workflow-level `defaults: run: shell: bash` in `docker-publish.yml`.**
+  Gitea Actions defaults each `run:` step to `sh` (dash), so every bash-syntax
+  step had to individually remember `shell: bash` — a discipline requirement
+  that failed twice (ed49b8d, b7197e8). Setting the default workflow-wide
+  eliminates the whole class. All pre-existing dash steps use only POSIX
+  syntax, so bash (a superset) runs them unchanged.
+
+### Fixed (CI)
+
+- **`promote-base-latest` now sets `shell: bash` on the base-latest re-tag
+  step.** The `b7197e8` fix (v1.2.4) moved the digest-compare into that step
+  with `set -euo pipefail`, but Gitea Actions' default step shell is `sh`
+  (dash), which rejects `-o pipefail` (`Illegal option -o pipefail`) and aborts
+  the step before the `crane copy` runs. On the v1.2.4 release (run 418) this
+  left `base-latest` un-promoted, still pointing at the v1.2.3 base — the four
+  consumer tags (`v1.2.4`, `latest`, `v1.2.4-studio`, `latest-studio`) were
+  unaffected because they `FROM` the exact `base-<hash>`, not `base-latest`.
+  Same footgun as `ed49b8d` (`resolve-versions needs shell: bash`).
+
+---
+
+## v1.2.4 — 2026-06-29
+
+Patch release. Headline: **pi `0.80.2` → `0.80.3`** (npm `latest`). Also ships a
+global gitignore baked into the image, secrets-via-`env_file`-only compose
+hardening, and a CI fix so `promote-base-latest` re-points `base-latest`
+reliably after a dry-run-first release. The mempalace pin stays `3.5.0`. The
+base image rebuilds because `Dockerfile.base` changed (the gitignore seed +
+`entrypoint-user.sh` wiring).
+
+### Added
+
+- **Global gitignore baked into the image.** A `~/.gitignore_global`
+  (`*.bak`, `*.bak.*`, `*~`, `*.orig`, `*.swp`, `*.tmp`) is seeded into the home
+  dir from `/etc/skel-devbox/` on first boot (seed-if-absent, like
+  `.bash_aliases`/`.inputrc`, so user edits survive recreate) and wired via
+  `git config --global core.excludesFile`. Personal/tooling backup artifacts are
+  now ignored across all repos in the container without per-repo `.gitignore`
+  entries. The `core.excludesFile` wiring is skipped if the user already set one.
+
+### Changed
+
+- **Secrets are now delivered to the container via `env_file: .env` only; the
+  `environment:` block no longer re-declares `GITEA_ACCESS_TOKEN`,
+  `GITEA_HOST`, or `GITHUB_PERSONAL_ACCESS_TOKEN`.** An `environment:` entry
+  both overrides `env_file:` and is interpolated from the host shell, so a
+  stale shell export (e.g. one auto-loaded by an opencode/dotenv hook) would
+  silently shadow the value in your `.env` — an updated token in `.env` never
+  reached the container. Delivering secrets via `env_file` only decouples the
+  container from whatever the host shell happens to export. No action needed:
+  `.env.example` already documents every supported variable. Affects
+  `docker-compose.yml` and the README “basic shape” snippet.
+
+### Fixed (CI)
+
+- **`promote-base-latest` now re-points `base-latest` reliably after a
+  dry-run-first release.** The job's gate previously required
+  `need_build == 'true'`, on the assumption that `need_build == false`
+  implied `base-latest` was already current. That assumption breaks when a
+  `workflow_dispatch` dry-run (`promote_latest=false`) pre-builds and pushes
+  `base-<hash>` first: the subsequent tag run then sees `need_build == false`
+  (probe hit) and **skipped** promotion, leaving `base-latest` pointing at the
+  *previous* base. (Observed 2026-06-27 releasing v1.2.3 via dry-run-then-tag
+  — `base-latest` ended up one base behind, lacking the mempalace self-heal.)
+  Now the gate runs on every tag release (or `promote_latest=true` dispatch),
+  and the no-op optimization moved **into** the step as a `crane digest`
+  compare: it re-tags only when `base-latest` actually differs from the
+  released `base-<hash>`, so genuine cache-hit releases stay a no-op while
+  stale aliases get corrected. No image-content change; base hash unaffected.
+
+---
+
+## v1.2.3 — 2026-06-27
+
+Patch release. Headline: **mempalace-mcp now self-heals** instead of latching
+`available=false` permanently after a slow cold-open. Also folds in the `yq`
+and mempalace-skill changes that were sitting unreleased. **No pi/mempalace
+version change** — pi npm `latest` is still `0.80.2` (= v1.2.2) and the
+mempalace pin stays `3.5.0`; the base image rebuilds purely because the
+`mempalace-toolkit` ref advances to pick up the self-heal extension.
+
+### Fixed
+
+- **mempalace-mcp self-heal — no more permanent `available=false` latch.**
+  The `mempalace.ts` pi extension (from `mempalace-toolkit`, bumped to
+  [`e12b624`](https://gitea.jordbo.se/joakimp/mempalace-toolkit/commit/e12b624))
+  previously tripped its per-request timeout on a slow virtiofs cold-open of
+  the palace, killed the child, and set `available=false` **forever** (no
+  respawn) — a pi restart was the only recovery.
+  - **Bounded respawn with capped exponential backoff** via `ensureAlive()`
+    (`MEMPALACE_MCP_MAX_RESPAWNS=2`, `MEMPALACE_MCP_RESPAWN_BACKOFF_MS=1000`;
+    set max to `0` to disable). Both `execute()` and initial startup route
+    through it. The respawn budget **resets on any successful JSON-RPC
+    response** (`onStdout`), so a healthy session can't slowly exhaust it.
+  - **Scoped init timeout** raised `120000 → 300000` ms (`MEMPALACE_MCP_INIT_TIMEOUT_MS`),
+    affecting **init only** — the per-call timeout stays `60000`
+    (`MEMPALACE_MCP_TIMEOUT_MS`) — so a genuine cold HNSW deserialize isn't
+    killed mid-open.
+  - **Concurrency hardening:** a generation counter prevents a late-exiting
+    killed process from clobbering a fresh respawn, and an explicit `healthy`
+    flag replaces the racy `proc != null` check.
+  - Note: the build-time `smoke-test.sh` verifies the extension is present and
+    deployed but does **not** exercise respawn behaviour — first live
+    validation is on a running container.
+- **`yq` is now mikefarah's Go yq, not Debian's Python `yq`.** The base image
+  previously apt-installed `yq`, which on Debian/Ubuntu is the unrelated
+  kislyuk/`yq` (a jq wrapper, v3.x) — incompatible with the mikefarah v4 syntax
+  the `cloud-init` repo's `provision.sh`/`deploy.sh` expect. Dropped the apt
+  package and install the mikefarah binary instead (multi-arch amd64/arm64,
+  following the repo's `latest` convention like `tealdeer`/`uv`; pin a tag
+  with `--build-arg YQ_VERSION=vX.Y.Z`). The build-time `smoke-test.sh` gate
+  asserts `yq --version` reports `mikefarah` **and** major **v4**, so both a
+  regression to the Python package and a surprise future yq v5 fail CI.
+
+### Changed
+
+- **Baked `mempalace` skill now teaches temporal grounding.** Added a
+  *Temporal grounding* rule to the image-baked
+  `skills/mempalace/SKILL.md` (Phase 1 wake-up + a matching anti-pattern):
+  before using relative time terms ("yesterday", "last week"), establish the
+  current date/time and compute the delta against the actual diary/drawer
+  timestamp. Explicitly calls out that a **container recreate or fresh session
+  is not a day boundary** — pi-devbox restarts several times a day, so two
+  entries minutes apart can straddle a recreate. Fixes agents mislabelling
+  same-day sessions as "yesterday".
+
+---
+
 ## v1.2.2 — 2026-06-24
 
 Patch release: pick up **pi `0.80.2`** (npm `latest`) and **mempalace `3.5.0`**,

Dockerfile.base

@@ -14,7 +14,7 @@
 # content-addressed over this file, so any byte change invalidates the
 # cache. Recommended cadence: once per release for security updates.
 #
-# BASE_REBUILD_DATE: 2026-06-09 (v1.0.0 — decoupled from opencode-devbox)
+# BASE_REBUILD_DATE: 2026-07-11 (Unreleased — typst PDF engine + xz-utils)
 #
 # ── Lineage note ─────────────────────────────────────────────────────
 # Adapted from opencode-devbox/Dockerfile.base (commit before v1.16.2).
@@ -46,16 +46,30 @@ ENV DEBIAN_FRONTEND=noninteractive
 # Additions vs the upstream opencode-devbox base (2026-06-09):
 #   pandoc       — Markdown↔HTML/PDF/etc. conversion. Required by pi-studio
 #                  preview/export pipelines and broadly useful for any
-#                  agent-driven document workflow. ~200 MB.
+#                  agent-driven document workflow. ~200 MB. NOTE: pandoc is
+#                  only the front-end — PDF output needs a back-end engine.
+#                  We ship `typst` (installed further down) as the
+#                  lightweight default engine (`pandoc --pdf-engine=typst`)
+#                  instead of a ~600 MB TeX Live install.
+#   xz-utils     — `xz` decompressor. tar shells out to it for `.tar.xz`
+#                  assets (typst ships .tar.xz). ~0.5 MB. Also generally
+#                  useful for extracting xz-compressed archives.
 #   graphviz     — `dot` rendering for many diagram tools. ~10 MB.
 #                  See the bundled `dot-watch` helper for live .dot -> PNG
 #                  re-render (handy with pi-studio's image preview).
 #   imagemagick  — image conversion / resizing for thumbnails, etc. ~50 MB.
-#   yq           — YAML-aware companion to jq.
+#   (yq is NOT apt-installed: Debian's `yq` is the unrelated Python tool;
+#    mikefarah's Go yq is installed as a pinned binary further down.)
 #   socat        — TCP relay. Powers `studio-expose`, which bridges
 #                  pi-studio's container-loopback server to the container's
 #                  external interface so a published port can reach it.
 #                  ~1 MB; generally useful for any port-forwarding need.
+#   nano         — small, non-modal terminal editor for users who don't want
+#                  a vi-based editor. ~2.8 MB installed; its deps (libc6,
+#                  libncursesw6, libtinfo6) are already pulled in by nvim/less/
+#                  htop/tmux, so it adds no extra packages. Companion to nvim
+#                  and the `micro` binary installed further down. EDITOR stays
+#                  nvim; users opt in via `export EDITOR=nano`.
 RUN apt-get update && \
     apt-get upgrade -y --no-install-recommends && \
     apt-get install -y --no-install-recommends \
@@ -66,7 +80,6 @@ RUN apt-get update && \
     openssh-client \
     gnupg \
     jq \
-    yq \
     ripgrep \
     fd-find \
     tree \
@@ -89,9 +102,11 @@ RUN apt-get update && \
     python3-pip \
     python3-venv \
     pandoc \
+    xz-utils \
     graphviz \
     imagemagick \
     socat \
+    nano \
     && ln -s /usr/bin/fdfind /usr/local/bin/fd \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/*
@@ -231,6 +246,33 @@ RUN ARCH=$(case "${TARGETARCH}" in amd64) echo "x86_64" ;; arm64) echo "arm64" ;
     ln -s /opt/nvim-linux-${ARCH}/bin/nvim /usr/local/bin/nvim && \
     nvim --version | head -1
 
+# micro — modern, non-modal terminal editor. Ships alongside nvim so users
+# who aren't comfortable with vi-style modal editing have a friendly option:
+# desktop-style keybindings (Ctrl+S save, Ctrl+Q quit, Ctrl+C/V/X, Ctrl+Z
+# undo), mouse support, and syntax highlighting out of the box. A single
+# static Go binary (~12 MB) installed from GitHub releases, exactly like
+# bat/eza/zoxide below. EDITOR stays nvim (see below); users opt in with
+# `export EDITOR=micro` or `git config --global core.editor micro`.
+#
+# NOTE: upstream moved zyedidia/micro -> micro-editor/micro. The old org URL
+# still 302s, but its /releases/latest redirect lands on ANOTHER /latest URL
+# (the org rename), so the tag-parsing idiom below would resolve "latest"
+# instead of a version. Use the canonical micro-editor/micro URL.
+# Arch asset naming differs from the others: amd64 -> linux64, arm64 ->
+# linux-arm64. The tarball extracts to micro-<version>/micro.
+ARG MICRO_VERSION=latest
+RUN ARCH=$(case "${TARGETARCH}" in amd64) echo "linux64" ;; arm64) echo "linux-arm64" ;; *) echo "linux64" ;; esac) && \
+    V="${MICRO_VERSION}" && \
+    if [ "$V" = "latest" ]; then \
+      V=$(curl -sI --retry 5 --retry-delay 5 --retry-all-errors "https://github.com/micro-editor/micro/releases/latest" | awk 'tolower($1)=="location:" { sub(/\r$/,"",$2); n=split($2,a,"/"); print a[n] }'); \
+    fi && \
+    V="${V#v}" && [ -n "$V" ] && \
+    echo "Installing micro ${V}" && \
+    curl -fsSL --retry 5 --retry-delay 5 --retry-all-errors "https://github.com/micro-editor/micro/releases/download/v${V}/micro-${V}-${ARCH}.tar.gz" | tar -xz -C /tmp && \
+    install /tmp/micro-${V}/micro /usr/local/bin/micro && \
+    rm -rf /tmp/micro-${V} && \
+    micro --version
+
 # bat
 ARG BAT_VERSION=latest
 RUN ARCH=$(case "${TARGETARCH}" in amd64) echo "x86_64" ;; arm64) echo "aarch64" ;; *) echo "x86_64" ;; esac) && \
@@ -289,15 +331,21 @@ RUN ARCH=$(case "${TARGETARCH}" in amd64) echo "x86_64" ;; arm64) echo "aarch64"
 # Always installed in the base. Set INSTALL_MEMPALACE=false at base-build
 # time to shave ~300 MB.
 #
-# Stall protection (fixed 2026-06-13): mempalace-mcp is launched by the
-# `mempalace.ts` pi extension from mempalace-toolkit (cloned below). That
-# extension now applies a per-REQUEST timeout in its JSON-RPC client and
-# kills the child on stall, so a virtiofs cold-open of chroma.sqlite3 /
-# HNSW load can no longer hang the pi TUI uninterruptibly. Tunables:
+# Stall protection (fixed 2026-06-13; self-heal added 2026-06-25):
+# mempalace-mcp is launched by the `mempalace.ts` pi extension from
+# mempalace-toolkit (cloned below). That extension applies a per-REQUEST
+# timeout in its JSON-RPC client and kills the child on stall, so a virtiofs
+# cold-open of chroma.sqlite3 / HNSW load can no longer hang the pi TUI
+# uninterruptibly. A stall-kill is no longer a permanent latch either: the
+# next tool call respawns the server with capped exponential backoff (the
+# budget resets on any successful response). Tunables:
 # MEMPALACE_MCP_TIMEOUT_MS (default 60000), MEMPALACE_MCP_INIT_TIMEOUT_MS
-# (default 120000); 0 disables. A standalone stdio-watchdog shim is NOT
-# needed — the extension already owns request/response correlation. See
-# CHANGELOG.md "Unreleased > Fixed".
+# (default 300000 — generous so a genuine first cold-open isn't killed),
+# MEMPALACE_MCP_MAX_RESPAWNS (default 2; 0 disables self-heal),
+# MEMPALACE_MCP_RESPAWN_BACKOFF_MS (default 1000); timeouts of 0 disable.
+# Defaults live in the extension, so no ENV is needed here. A standalone
+# stdio-watchdog shim is NOT needed — the extension already owns
+# request/response correlation. See CHANGELOG.md "Unreleased > Fixed".
 ARG INSTALL_MEMPALACE=true
 # Pin to a known-good version. Bump deliberately, not implicitly: an
 # unpinned install silently swept in mempalace 3.3.x/3.4.0 with a broken
@@ -404,6 +452,52 @@ RUN ARCH=$(case "${TARGETARCH}" in amd64) echo "x86_64" ;; arm64) echo "aarch64"
     chmod +x /usr/local/bin/tldr && \
     tldr --version
 
+# ── typst — lightweight PDF engine for pandoc (Markdown→PDF) ─────────
+# pandoc (apt-installed above) is only a front-end; rendering PDF needs a
+# back-end engine. Rather than a ~600 MB TeX Live install, we ship typst:
+# a single ~30 MB static Rust binary with no LaTeX dependency. pi-studio's
+# PDF export (studio_export_pdf) and pandoc invocations use it via
+# `pandoc --pdf-engine=typst`. A fuller TeX Live remains the higher-
+# fidelity fallback for anyone who needs LaTeX-exact output (not shipped
+# here — install on demand or in a future variant).
+#
+# Follows the `latest` GitHub-release convention (like tealdeer/uv/bat).
+# typst ships a `.tar.xz` asset (hence xz-utils in the apt layer above)
+# that extracts to typst-<arch>-unknown-linux-musl/typst. Pin a specific
+# tag with --build-arg TYPST_VERSION=vX.Y.Z.
+ARG TYPST_VERSION=latest
+RUN ARCH=$(case "${TARGETARCH}" in amd64) echo "x86_64" ;; arm64) echo "aarch64" ;; *) echo "x86_64" ;; esac) && \
+    V="${TYPST_VERSION}" && \
+    if [ "$V" = "latest" ]; then \
+      V=$(curl -sI --retry 5 --retry-delay 5 --retry-all-errors "https://github.com/typst/typst/releases/latest" | awk 'tolower($1)=="location:" { sub(/\r$/,"",$2); n=split($2,a,"/"); print a[n] }'); \
+    fi && \
+    V="${V#v}" && [ -n "$V" ] && \
+    echo "Installing typst ${V}" && \
+    curl -fsSL --retry 5 --retry-delay 5 --retry-all-errors "https://github.com/typst/typst/releases/download/v${V}/typst-${ARCH}-unknown-linux-musl.tar.xz" | tar -xJ -C /tmp && \
+    install /tmp/typst-${ARCH}-unknown-linux-musl/typst /usr/local/bin/typst && \
+    rm -rf /tmp/typst-${ARCH}-unknown-linux-musl && \
+    typst --version
+
+# ── yq (mikefarah) — YAML processor, jq's companion for YAML ─────────
+# Installed as the mikefarah Go binary — NOT Debian's `yq` apt package, which
+# is the unrelated Python kislyuk/yq (a jq wrapper with different syntax and
+# version line, e.g. 3.x). The cloud-init repo's deploy.sh/provision.sh
+# require mikefarah yq v4 (the unrelated Debian python yq is v3.x). Follows
+# the repo's `latest` convention (like tealdeer/uv/etc.); the smoke test pins
+# the contract to major v4, so a future yq v5 fails CI instead of silently
+# breaking provision.sh. Pin a specific tag with --build-arg YQ_VERSION=vX.Y.Z.
+ARG YQ_VERSION=latest
+RUN ARCH=$(case "${TARGETARCH}" in amd64) echo "amd64" ;; arm64) echo "arm64" ;; *) echo "amd64" ;; esac) && \
+    V="${YQ_VERSION}" && \
+    if [ "$V" = "latest" ]; then \
+      V=$(curl -sI --retry 5 --retry-delay 5 --retry-all-errors "https://github.com/mikefarah/yq/releases/latest" | awk 'tolower($1)=="location:" { sub(/\r$/,"",$2); n=split($2,a,"/"); print a[n] }'); \
+    fi && \
+    [ -n "$V" ] && \
+    echo "Installing mikefarah yq ${V}" && \
+    curl -fsSL --retry 5 --retry-delay 5 --retry-all-errors "https://github.com/mikefarah/yq/releases/download/${V}/yq_linux_${ARCH}" -o /usr/local/bin/yq && \
+    chmod +x /usr/local/bin/yq && \
+    yq --version
+
 # ── AWS CLI v2 (for SSO/Bedrock authentication) ─────────────────────
 RUN ARCH=$(case "${TARGETARCH}" in \
       amd64) echo "x86_64" ;; \
@@ -451,8 +545,9 @@ ENV PATH="/home/${USER_NAME}/.pi/npm-global/bin:${PATH}"
 
 # ── Shell defaults (bash history, aliases, readline) ─────────────────
 RUN mkdir -p /etc/skel-devbox
-COPY rootfs/home/developer/.bash_aliases /etc/skel-devbox/.bash_aliases
-COPY rootfs/home/developer/.inputrc      /etc/skel-devbox/.inputrc
+COPY rootfs/home/developer/.bash_aliases     /etc/skel-devbox/.bash_aliases
+COPY rootfs/home/developer/.inputrc          /etc/skel-devbox/.inputrc
+COPY rootfs/home/developer/.gitignore_global /etc/skel-devbox/.gitignore_global
 
 # ── Entrypoint ────────────────────────────────────────────────────────
 COPY rootfs/usr/local/lib/pi-devbox/ /usr/local/lib/pi-devbox/

README.md

@@ -29,13 +29,18 @@ on the host.
 - ChromaDB embedding model pre-warmed at build time (`all-MiniLM-L6-v2`)
 
 The host-mounted palace at `~/.mempalace` is shared across the host and
-this container so all your agents share one brain.
+this container so all your agents share one brain. To instead share a palace
+across *several* containers/harnesses, set `MEMPALACE_REMOTE_URL` to a shared
+MemPalace HTTP endpoint (see `.env.example` and `docker-compose.mempalace.yml`);
+the bridge then connects over HTTP and spawns no local server.
 
 ### Modern CLI tooling
 
 | Tool | Purpose |
 |---|---|
-| `nvim` | Neovim text editor |
+| `nvim` | Neovim text editor (modal / vi-style) |
+| `nano` | Small non-modal editor (on-screen shortcut hints) |
+| `micro` | Modern non-modal editor (Ctrl+S/Ctrl+Q keys, mouse, syntax highlighting) |
 | `tmux` | Terminal multiplexer (configured for 0-indexed sessions) |
 | `ripgrep`, `fd` | Fast file content / filename search |
 | `fzf` | Fuzzy finder |
@@ -49,6 +54,11 @@ this container so all your agents share one brain.
 | `gosu` | Privilege de-escalation in entrypoint |
 | `htop`, `tree`, `less` | Inspection utilities |
 
+The default `$EDITOR` is `nvim`. Three editors ship so you can pick your
+comfort level — if you'd rather not use a vi-style editor, `nano` and `micro`
+are both non-modal. Set your preference with `export EDITOR=micro` (or `nano`)
+in your shell profile, and/or `git config --global core.editor micro`.
+
 ### Document and image tooling
 
 - `pandoc` — universal Markdown↔HTML/Org/RST/etc. converter
@@ -139,8 +149,10 @@ Currently published:
 
 Planned for an upcoming minor release:
 
-- `joakimp/pi-devbox:latest-studio-tex` — `-studio` plus `texlive-xetex`
-  for PDF export from Studio. Adds ~600 MB on top of `-studio`.
+- *(shipped in Unreleased/base)* **PDF export from Studio/pandoc** now works:
+  the base image ships **`typst`** as the PDF engine (`pandoc --pdf-engine=typst`),
+  a single ~30 MB static binary — no separate `-tex` variant needed.
+  `texlive-xetex` stays the higher-fidelity fallback (install on demand).
 
 ## Using pi-studio (`-studio` variant)
 
@@ -275,9 +287,10 @@ Assuming the compose file publishes `127.0.0.1:8765:8765` (see method B):
 > until step 2 runs. If the browser can't connect, verify Studio is up
 > (`/studio --status`) and the bridge is running (`ps aux | grep socat`).
 
-> PDF export (`/studio-pdf`, `studio_export_pdf`) needs a LaTeX engine,
-> which is **not** in `-studio` (only the planned `-studio-tex`). HTML
-> export, KaTeX, Mermaid, and all REPL features work without it.
+> PDF export (`/studio-pdf`, `studio_export_pdf`) uses **`typst`**, shipped in
+> the base image as the pandoc PDF engine (`pandoc --pdf-engine=typst`). For
+> LaTeX-exact output you can install `texlive-xetex` on demand as a heavier
+> fallback. HTML export, KaTeX, Mermaid, and all REPL features work regardless.
 
 ### Graphviz diagrams in Studio: `dot-watch`
 
@@ -319,9 +332,10 @@ services:
     environment:
       - TERM=xterm-256color
       # - STUDIO_EXPOSE=1          # -studio only: auto-start the socat bridge on boot
-      - GITEA_ACCESS_TOKEN=${GITEA_ACCESS_TOKEN:-}
-      - GITEA_HOST=${GITEA_HOST:-}
-      - GITHUB_PERSONAL_ACCESS_TOKEN=${GITHUB_PERSONAL_ACCESS_TOKEN:-}
+      # Secrets (GITEA_*, GITHUB_*, …) come from env_file: .env above — not
+      # duplicated here. An environment: entry overrides env_file and is
+      # interpolated from the host shell, so a stale shell export would
+      # silently shadow your .env. See .env.example for the full list.
     volumes:
       # Workspace: your host source tree
       - ${WORKSPACE_PATH:-.}:/workspace

docker-compose.mempalace.yml

@@ -0,0 +1,84 @@
+# Shared MemPalace server (optional) — one palace for many clients.
+#
+# Runs `mempalace-mcp` over HTTP so several containers/harnesses (pi +
+# opencode + native) can share ONE palace instead of each keeping its own.
+# Point every client at it by setting, in that client's .env:
+#
+#   MEMPALACE_REMOTE_URL=http://<reachable-host>:8765/mcp
+#
+# (see .env.example). When set, the client connects over HTTP and does NOT
+# spawn its own local mempalace-mcp.
+#
+# Start:   docker compose -f docker-compose.mempalace.yml up -d
+# Stop:    docker compose -f docker-compose.mempalace.yml down
+# Logs:    docker compose -f docker-compose.mempalace.yml logs -f
+#
+# Why reuse the devbox image? mempalace-mcp is already installed in it, and
+# reusing it GUARANTEES the server's mempalace version matches the clients'
+# (both are pinned by the same image build). Override with a slimmer image via
+# MEMPALACE_SERVER_IMAGE if you prefer (it must provide `mempalace-mcp`).
+#
+# ⚠ SECURITY: mempalace-mcp's HTTP transport has NO authentication of its own.
+# Do NOT expose port 8765 to an untrusted network. The default below binds to
+# 127.0.0.1 (host loopback) only. To let sibling containers reach it, either
+# attach them to the shared `mempalace-net` network (container-to-container, no
+# host port needed — use http://mempalace-server:8765/mcp), or front it with a
+# reverse proxy that enforces MEMPALACE_REMOTE_TOKEN as `Authorization: Bearer`.
+
+name: mempalace-server
+
+services:
+  mempalace:
+    image: ${MEMPALACE_SERVER_IMAGE:-joakimp/pi-devbox:latest}
+    container_name: mempalace-server
+    # Bypass the devbox entrypoint (dev-shell/LAN/config setup) and run the
+    # HTTP MCP server directly. HOME + explicit --palace pin the data path so
+    # it does not depend on the image's default user/HOME. Runs as root so it
+    # can initialise the fresh named volume; the volume is dedicated to this
+    # server (clients reach it over HTTP, never by mounting it).
+    entrypoint: []
+    user: "0:0"
+    environment:
+      - HOME=/data
+    command:
+      - mempalace-mcp
+      - --transport
+      - http
+      - --host
+      - "0.0.0.0"
+      - --port
+      - "8765"
+      - --palace
+      - /data/.mempalace
+    restart: unless-stopped
+    # Loopback-only by default (see SECURITY note). Use "8765:8765" to expose on
+    # all host interfaces, or drop `ports:` entirely and rely on mempalace-net.
+    ports:
+      - "127.0.0.1:8765:8765"
+    volumes:
+      # The shared palace data — precious; back this up.
+      - mempalace-shared:/data/.mempalace
+      # Embedding-model cache (~79 MB, disposable) so search does not re-download.
+      - mempalace-shared-chroma:/data/.cache/chroma
+    networks:
+      - mempalace-net
+    healthcheck:
+      # A tools/list round-trip proves the server is answering MCP (python3 is
+      # always present — mempalace itself is a python tool in the image).
+      test:
+        - CMD
+        - python3
+        - -c
+        - "import urllib.request,json; d=json.dumps({'jsonrpc':'2.0','id':1,'method':'tools/list','params':{}}).encode(); r=urllib.request.Request('http://127.0.0.1:8765/mcp',data=d,headers={'Content-Type':'application/json','Accept':'application/json'}); urllib.request.urlopen(r,timeout=5).read()"
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 60s
+
+volumes:
+  mempalace-shared:
+  mempalace-shared-chroma:
+
+networks:
+  mempalace-net:
+    name: mempalace-net

docker-compose.yml

@@ -31,9 +31,13 @@ services:
       - .env
     environment:
       - TERM=xterm-256color
-      - GITEA_ACCESS_TOKEN=${GITEA_ACCESS_TOKEN:-}
-      - GITEA_HOST=${GITEA_HOST:-}
-      - GITHUB_PERSONAL_ACCESS_TOKEN=${GITHUB_PERSONAL_ACCESS_TOKEN:-}
+      # Secrets (GITEA_*, GITHUB_*, and any others) are delivered to the
+      # container via `env_file: .env` above — do NOT duplicate them here.
+      # An `environment:` entry overrides env_file AND is interpolated from
+      # the host shell, so a stale shell export (e.g. one auto-loaded by a
+      # dotenv hook) would silently shadow the value in your .env. Keeping
+      # secrets env_file-only decouples the container from the host shell.
+      # See .env.example for the full list of supported variables.
     volumes:
       # Host workspace — mount your project here
       - ${WORKSPACE_PATH:-.}:/workspace
@@ -72,7 +76,10 @@ services:
       # Persist uv data (Python installs, tool installs)
       - devbox-uv:/home/developer/.local/share/uv
 
-      # Optional: persist MemPalace data (conversation memory, knowledge graph)
+      # Optional: persist MemPalace data (conversation memory, knowledge graph).
+      # Applies to the LOCAL palace only (the default). In EXTERNAL mode
+      # (MEMPALACE_REMOTE_URL set in .env) the shared server owns the data, so
+      # this volume is irrelevant.
       # - devbox-palace:/home/developer/.mempalace
 
       # Optional: persist ChromaDB embedding model cache (~79 MB)

entrypoint-user.sh

@@ -33,7 +33,7 @@ fi
 # directly.
 SKEL_DIR="/etc/skel-devbox"
 if [ -d "$SKEL_DIR" ]; then
-  for f in .bash_aliases .inputrc; do
+  for f in .bash_aliases .inputrc .gitignore_global; do
     if [ -f "$SKEL_DIR/$f" ] && [ ! -e "$HOME/$f" ]; then
       cp "$SKEL_DIR/$f" "$HOME/$f"
     fi
@@ -91,6 +91,12 @@ fi
 if [ -n "${GIT_USER_EMAIL:-}" ] && ! git config --global user.email &>/dev/null; then
   git config --global user.email "$GIT_USER_EMAIL"
 fi
+# Global gitignore for personal/tooling artifacts (*.bak, *~, *.orig, ...).
+# Seeded above into $HOME/.gitignore_global from /etc/skel-devbox. Point git at
+# it only if the user has not already set their own core.excludesFile.
+if [ -f "$HOME/.gitignore_global" ] && ! git config --global core.excludesFile &>/dev/null; then
+  git config --global core.excludesFile "$HOME/.gitignore_global"
+fi
 
 # ── pi: deploy toolkit + extensions + mempalace bridge ─────────────
 # pi is always installed in pi-devbox; no INSTALL_PI guard needed.

rootfs/home/developer/.bash_aliases

@@ -54,6 +54,38 @@ alias gs='git status'
 alias gd='git diff'
 alias gl='git log --oneline --graph --decorate -20'
 
+# ── Host SSH reachability check (once per container lifetime) ───────────────
+# Warns at first shell startup if the Mac host is not reachable via SSH.
+# Only runs inside a container, only if the jump key exists, and only once
+# per container lifetime (/tmp flag is cleared on recreate).
+_devbox_check_host_ssh() {
+  [ -f "/.dockerenv" ] || return 0
+  local ssh_cfg="$HOME/.ssh-local/config"
+  [ -f "$ssh_cfg" ] || return 0
+  local key_pub="$HOME/.ssh-local/devbox_jump_ed25519.pub"
+  [ -f "$key_pub" ] || return 0
+  local flag="/tmp/.devbox_host_ssh_ok"
+  [ -f "$flag" ] && return 0
+  if ssh -F "$ssh_cfg" \
+        -o BatchMode=yes \
+        -o ConnectTimeout=2 \
+        -o StrictHostKeyChecking=accept-new \
+        mac true 2>/dev/null; then
+    touch "$flag"
+    return 0
+  fi
+  local pub_key
+  pub_key=$(cat "$key_pub")
+  printf '\n\033[1;33m⚠  devbox: Mac host not reachable via SSH\033[0m\n'
+  printf '   Some tools use SSH to run commands on the Mac host.\n'
+  printf '   Fix (run both on the Mac):\n\n'
+  printf '   \033[1mStep 1\033[0m  System Settings → General → Sharing → Remote Login → ON\n\n'
+  printf '   \033[1mStep 2\033[0m  echo '"'"'%s'"'"' >> ~/.ssh/authorized_keys\n' "$pub_key"
+  printf '\n   Then open a new shell in the container to verify.\n\n'
+}
+_devbox_check_host_ssh
+unset -f _devbox_check_host_ssh
+
 # ── LAN access via the host (dssh) ───────────────────────────────────
 # When running on a VM-backed host (macOS OrbStack / Docker Desktop), the
 # entrypoint's setup-lan-access.sh generates ~/.ssh-local/config so the host

rootfs/home/developer/.gitignore_global

@@ -0,0 +1,11 @@
+# Global gitignore — personal/tooling artifacts (applies to all repos in the container)
+# Seeded into $HOME/.gitignore_global by entrypoint-user.sh and wired via
+# `git config --global core.excludesFile`. Edit freely; it is yours after first boot.
+
+# backup / editor / merge artifacts
+*.bak
+*.bak.*
+*~
+*.orig
+*.swp
+*.tmp

rootfs/usr/local/share/pi-devbox/skills/mempalace/SKILL.md

@@ -43,6 +43,29 @@ Run these immediately when a session begins, before responding to the user:
 
 Do NOT announce this to the user. Just do it silently to orient yourself.
 
+### Temporal grounding — compute time deltas, don't guess
+
+Diary entries and drawers carry real timestamps (`timestamp`, `created_at`).
+Before describing *when* something happened — "yesterday", "earlier today",
+"last week", "a while back" — **establish the current date/time first and
+compute the delta against the actual timestamp.** Get "now" from the injected
+session date or by running `date` in a shell; never infer it.
+
+**A container recreate or a fresh session is NOT a day boundary.** A pi-devbox
+container is frequently restarted — often several times within the *same* day —
+and each restart begins a new session with a fresh wake-up. Do not reason "new
+session ⇒ last session was yesterday": two diary entries 90 minutes apart can
+straddle a container recreate. The only authoritative clock is the timestamp on
+the memory, not the session/container boundary.
+
+**Practical rule:** prefer explicit, checkable phrasing — e.g. "earlier today,
+~8h ago (both 2026-06-25)" — over a vague relative term. If you catch yourself
+about to write "yesterday" / "last week", subtract `now − entry.timestamp` and
+state the computed result. (Remember timestamps may be UTC while the wall clock
+is local — reconcile the offset before computing the delta.) Note too that
+session feeders can lag up to a week (see *Multi-harness palace*), so a recent
+absence in `wing_conversations` is not proof nothing happened.
+
 ### Phase 2: Active Session (during work)
 
 #### Search Before You Speak
@@ -294,6 +317,7 @@ Entity-relationship triples with temporal validity. Query with `mempalace_kg_que
 ## Anti-Patterns
 
 - **Don't guess when you can search.** If a question touches past work, search first.
+- **Don't infer elapsed time from session or container boundaries.** A restart isn't a new day. Compare the actual timestamp (`timestamp` / `created_at`) against the current date/time before saying "yesterday", "last week", etc.
 - **Don't skip the diary.** A session without a diary entry is a session forgotten.
 - **Don't summarize drawer content.** File verbatim — the embedding model needs the original words.
 - **Don't mine .git directories or node_modules.** The CLI miner respects .gitignore by default.

scripts/check-workflow-shell.sh

@@ -0,0 +1,65 @@
+#!/usr/bin/env bash
+# Gitea-accurate guard against the recurring "bash syntax under the default
+# sh/dash shell" footgun (ed49b8d resolve-versions; b7197e8 promote-base-latest,
+# run 418).
+#
+# WHY A CUSTOM CHECK AND NOT JUST actionlint:
+#   actionlint models *GitHub* Actions, whose default `run` shell is bash. It
+#   therefore assumes a step that omits `shell:` runs under bash, and does NOT
+#   flag `set -o pipefail` there. Gitea Actions' default is `sh` (dash), so the
+#   exact bug we hit (omit `shell:`, use bash syntax) is invisible to actionlint.
+#   actionlint only fires when a step *explicitly* declares `shell: sh`.
+#
+# THE INVARIANT THIS ENFORCES:
+#   Every `run:` step in every .gitea/workflows/*.yml must resolve to an
+#   effective shell of `bash` — via the step's own `shell:`, a job-level
+#   `defaults.run.shell`, or a workflow-level `defaults.run.shell`. Any step
+#   that would fall through to Gitea's `sh` default is a FAILURE, because a
+#   future author adding bash syntax to it fails silently in CI.
+#
+# Pair this with actionlint (which catches explicit `shell: sh` + bash syntax,
+# expression errors, and much else). Together they cover the class on Gitea.
+set -euo pipefail
+
+WF_DIR="${1:-.gitea/workflows}"
+
+python3 - "$WF_DIR" <<'PY'
+import sys, glob, os
+try:
+    import yaml
+except ImportError:
+    sys.stderr.write("ERROR: python3 yaml module missing (apt install python3-yaml)\n")
+    sys.exit(2)
+
+wf_dir = sys.argv[1]
+files = sorted(glob.glob(os.path.join(wf_dir, "*.yml")) + glob.glob(os.path.join(wf_dir, "*.yaml")))
+if not files:
+    sys.stderr.write(f"ERROR: no workflow files under {wf_dir}\n")
+    sys.exit(2)
+
+problems = []
+for f in files:
+    with open(f) as fh:
+        doc = yaml.safe_load(fh) or {}
+    wf_shell = (((doc.get("defaults") or {}).get("run") or {}).get("shell"))
+    jobs = doc.get("jobs") or {}
+    for jname, job in jobs.items():
+        job = job or {}
+        job_shell = (((job.get("defaults") or {}).get("run") or {}).get("shell"))
+        steps = job.get("steps") or []
+        for i, step in enumerate(steps):
+            step = step or {}
+            if "run" not in step:
+                continue  # `uses:` steps have no shell
+            eff = step.get("shell") or job_shell or wf_shell or "sh"  # Gitea default = sh
+            if eff != "bash":
+                name = step.get("name") or f"step[{i}]"
+                problems.append(f"{f}: job '{jname}' / '{name}': effective shell = '{eff}' (Gitea default is sh; declare shell: bash or a bash default)")
+
+if problems:
+    sys.stderr.write("Workflow shell guard FAILED — bash default not guaranteed:\n")
+    for p in problems:
+        sys.stderr.write(f"  - {p}\n")
+    sys.exit(1)
+print(f"Workflow shell guard OK — all run: steps in {len(files)} workflow file(s) resolve to bash.")
+PY

scripts/smoke-test.sh

@@ -6,6 +6,8 @@
 # Verifies:
 #   - pi binary present and (if EXPECTED_PI_VERSION set) matches CI's resolved version
 #   - new v1.0.0 base additions (pandoc, graphviz, imagemagick, yq, tealdeer)
+#   - typst PDF engine for pandoc (Unreleased) — `pandoc --pdf-engine=typst`
+#   - non-modal editors nano + micro (alongside nvim)
 #   - tmux 0-indexing baked in /etc/tmux.conf (required for pi-studio variants)
 #   - pi-toolkit cloned at /opt/pi-toolkit
 #   - pi-extensions cloned at /opt/pi-extensions
@@ -24,7 +26,8 @@ set -euo pipefail
 IMAGE="${1:?usage: $0 <image>}"
 PASS=0; FAIL=0
 # pi-devbox v1.0.0 (decoupled from opencode-devbox) added pandoc, graphviz,
-# imagemagick, yq, tealdeer, and a baked /etc/tmux.conf. Local arm64 build
+# imagemagick, yq, tealdeer, a baked /etc/tmux.conf, and the non-modal
+# editors nano + micro (~15 MB combined). Local arm64 build
 # observed 3.20 GB. CI amd64 builds may differ slightly; threshold below
 # carries +300 MB margin to absorb arch differences without false reds.
 # Tighten in a follow-up release once amd64 actuals are observed in CI logs.
@@ -71,12 +74,16 @@ run "git"             "git --version"
 run "aws"             "aws --version"
 run "uv"              "uv --version"
 run "nvim"            "nvim --version"
+run "nano"            "nano --version"
+run "micro"           "micro --version"
 run "mempalace-mcp"   "mempalace-mcp --help"
 # v1.0.0 base additions — verify presence and basic functionality.
 run "pandoc"          "pandoc --version"
+run "typst"           "typst --version"
+run "pandoc+typst PDF engine" "printf '# hi\n' | pandoc --pdf-engine=typst -o /tmp/_smoke.pdf - && test -s /tmp/_smoke.pdf; rm -f /tmp/_smoke.pdf"
 run "graphviz (dot)"  "dot -V"
 run "imagemagick"     "magick --version"
-run "yq"              "yq --version"
+run "yq (mikefarah v4)" "yq --version | grep -qE 'mikefarah.*version v4'"
 run "tldr (tealdeer)" "tldr --version"
 run "socat"           "socat -V"
 run "studio-expose helper" "test -x /usr/local/bin/studio-expose"