complete: board-github-data-seam (PyAutoBrain#303)

active.md

@@ -1,21 +1,5 @@
 # Active Tasks
 
-## board-github-data-seam
-- issue: https://github.com/PyAutoLabs/PyAutoBrain/issues/303 (issued 2026-08-27)
-- issued: 2026-08-27
-- prompt: active/board_without_gh_phase1_seam.md
-- status: library-dev — not started; routed by /start_dev, no worktree claimed yet
-  (this session is remote: no ~/Code/PyAutoLabs-wt, so /start_library runs on the
-  machine that picks the task up).
-- worktree: ~/Code/PyAutoLabs-wt/board-github-data-seam/ (to be created by /start_library)
-- repos: PyAutoBrain
-- phase: 1 of 2 — phase 2 is draft/feature/pyautobrain/board_without_gh_phase2_legs.md,
-  blocked-by this one; parent design + probe evidence in
-  draft/feature/pyautobrain/board_without_gh.md
-- why: the board is the morning door and eleven legs are dark on the surface it is
-  read from. The 2026-08-27 probe closed the two cheaper options — $GH_TOKEN 403s
-  every repo path — so the injection seam is the design.
-
 ## transformed-message-factor-gradient-unpack
 - issue: https://github.com/PyAutoLabs/PyAutoFit/issues/1501 (issued 2026-08-19)
 - issued: 2026-08-19

complete/2026/08/board-github-data-seam.md

@@ -0,0 +1,157 @@
+- completed: 2026-08-27
+- issue: https://github.com/PyAutoLabs/PyAutoBrain/issues/303 (closed)
+- prs:
+  - https://github.com/PyAutoLabs/PyAutoBrain/pull/305 (merged, `ef887f8`)
+- classification: feature (PyAutoBrain; phase 1 of 2)
+- summary: |
+    The Brain board is the morning door, and on the surface it is actually read
+    from — a session on a phone — eleven of its legs were dark: they read GitHub
+    through `gh api`, and a remote session has no `gh`. Phase 1 gives the
+    renderer a way to be handed what it cannot fetch, and proves it on the seven
+    overnight rows.
+
+## The shape, and why it is a seam
+
+`mcp__github__*` is an **agent** capability. `board/_board.py` is a subprocess
+and cannot reach it however it is invoked — which is why "port the board to
+MCP" was never a substitution and always a seam:
+
+```
+pyauto-brain board --github-data <file>      # {endpoint: response}
+```
+
+Keyed by the endpoint string exactly as `gh_json()` receives it, so injected and
+live data are interchangeable by construction rather than by convention: a key
+that drifts from its call site simply misses, and a miss is a state the board
+already renders honestly. The `gh` path is untouched — no flag, no injected map,
+identical behaviour on a developer box.
+
+The `/board` skill (which *is* the agent) gathers when `command -v gh` fails,
+writes the file, and invokes the renderer. The file's shape is a documented
+contract in `board/AGENTS.md`, because it outlives whoever wrote it.
+
+## The three rules, all one invariant
+
+Restatements of what the 2026-08-26 degraded-render work established:
+
+1. A **miss is *could not ask*** (`None`) — never an empty answer. That is
+   exactly how a board goes green by never looking.
+2. An explicit **`null` means the gatherer's own fetch failed** — same outcome,
+   stated rather than implied. `{"workflow_runs": []}` is the opposite: a real
+   answer that happens to be empty.
+3. A **malformed or missing file is fatal, not degrading.** Every leg reading
+   `could not read` renders as a GitHub outage; if the truth is that the
+   gatherer wrote a broken file, the board must say so instead.
+
+## Key traps
+
+- **The contract includes the field names.** Store each response as GitHub
+  returns it. The overnight row's age reads `created_at`; the first
+  hand-written fixture used `updated_at`, and the row rendered `success (?)` —
+  which reads as a rendering bug rather than a bad input file. Documented in
+  both places and asserted in the tests, because the next gatherer will guess
+  the same way.
+
+- **Phase 0 was a probe, not a design.** The prompt offered three options and
+  said to probe before building; the probe (2026-08-27) returned 200 on `/user`
+  and `/rate_limit` and **403 on every repo-scoped path**, which killed options
+  2 and 3 and selected this one. Recorded in the parent prompt so the next
+  session starts at the design. If an org admin ever connects the Claude GitHub
+  App, `gh_json()` becomes a plain REST helper on every surface and this seam is
+  deletable — which is why it is small.
+
+- **Proved on the surface it exists for.** The end-to-end check ran in this
+  session, which has no `gh`: one injected endpoint rendered
+  `✓ PyAutoLabs/PyAutoBrain/nightly-release.yml — success (6h)` with its run
+  link while the six uninjected legs still said `could not read`, and the render
+  stayed degraded and non-green.
+
+- **PyAutoMind's PR-open event fired no workflows — twice, in one session.**
+  Both close-out PRs (#354, #357) showed **zero** check runs at open, on diffs
+  matching three path filters plus `spawn_drift.yml`, which has none. Actions
+  was healthy throughout: other PRs on the repo ran normally in the same
+  minutes, and a `workflow_dispatch` started instantly. #354 recovered the
+  moment an unrelated push produced a `synchronize` event, which fired all four
+  at once.
+
+  The 2026-08-26 record logged this as a one-off GitHub-side miss; at two
+  occurrences it is worth treating as a repeatable condition on this repo
+  rather than a blip, and the response is the one that record named: **verify
+  by dispatch, never merge unchecked**. Note that only `firewall_gate.yml` is
+  safely dispatchable here — `lifecycle_drift.yml` and `dashboard_refresh.yml`
+  self-heal `main` on a manual dispatch (they reset to `origin/main` and push),
+  so dispatching them proves nothing about the branch and writes to main.
+  Where they cannot be dispatched, run their commands locally — they are the
+  same four: `lifecycle.py check`, `lifecycle.py index --check`,
+  `registry_toc.py --check`, `intake dashboard --check`.
+
+## Validation
+
+569 PyAutoBrain tests (`-n auto`), `ruff` clean. 8 new tests, fixture-driven,
+with `BOARD_GH` pointed at a nonexistent command so a leak to a live `gh` fails
+loudly rather than answering.
+
+## Scope note
+
+A full seven-row live render was not demonstrable from the shipping session:
+three of the watched repos (PyAutoHeart, PyAutoHands, autolens_assistant) were
+outside its repository scope, so the gatherer could not reach them. The
+mechanism is identical per leg.
+
+## Follow-up
+
+`draft/feature/pyautobrain/board_without_gh_phase2_legs.md` — the remaining four
+legs (versions, community, resume, upkeep), now unblocked. Two of them live in
+scripts rather than in `_board.py`, so the phase's real question is whether each
+script grows its own seam or the board gathers for them; answer it in this
+contract's terms rather than inventing a second shape.
+
+## Original prompt
+
+# Board phase 1: the injection seam, proven on the overnight legs
+
+Type: feature
+Target: pyautobrain
+Repos:
+- PyAutoBrain
+Difficulty: medium
+Autonomy: supervised
+Priority: normal
+Status: formalised
+Filed: 2026-08-27
+Issued: 2026-08-27
+Parent: draft/feature/pyautobrain/board_without_gh.md
+
+Phase 1 of the board's gh-less work. The parent prompt holds the design, the
+measurements and the 2026-08-27 probe that selected this option; read it first
+and do not re-derive them.
+
+## Scope
+
+Build the seam and prove it on **one** leg family — the seven `overnight:
+could not read <repo>/<workflow>` rows, the largest block of the eleven.
+
+- `board/_board.py` accepts pre-fetched GitHub JSON (`--github-data <file>`),
+  and `gh_json()` reads from it when present. The `gh` path is untouched, so a
+  dev box behaves exactly as today.
+- The `/board` skill — which *is* the agent, and so is the only thing that can
+  call `mcp__github__*` — gathers that JSON when `command -v gh` fails, writes
+  it, and invokes the renderer with the flag. `_board.py` stays a pure
+  renderer; a subprocess cannot reach MCP tools and this is the whole reason
+  the seam exists.
+- The file's shape is a documented contract in `board/AGENTS.md` (extend
+  "Reading the board in a remote session"), because it outlives whoever writes
+  it.
+
+## Guards
+
+- Tests drive the seam from a **fixture file**, never a live call.
+- A leg with no injected data still reports `could not read` — the invariant
+  from the 2026-08-26 work: no leg may substitute an empty answer for an
+  unasked question. A green badge still requires a render that read everything.
+
+## Done when
+
+The overnight rows are live in a remote render, the degraded count drops by
+seven, the dev-box path is byte-identical, and `tests/test_board_degraded.py`
+plus the new fixture tests pass.

complete/index.md

@@ -6,7 +6,7 @@ Token-light navigation over the finished-work records (schema:
 only then grep a dated bucket. Curators: edit the band between the CURATED
 markers; everything below GENERATED is rebuilt.
 
-1163 records across 7 buckets.
+1164 records across 7 buckets.
 
 <!-- CURATED:START -->
 ## Highlights
@@ -42,6 +42,7 @@ _(curate hard-won records here — survives regeneration.)_
 - [autolens-tracer-validation-guards](2026/08/autolens-tracer-validation-guards.md)
 - [autoreduce-pypi-name-collision](2026/08/autoreduce-pypi-name-collision.md)
 - [blackjax-16-upgrade-smoke](2026/08/blackjax-16-upgrade-smoke.md)
+- [board-github-data-seam](2026/08/board-github-data-seam.md)
 - [border-relocator-backend-parity](2026/08/border-relocator-backend-parity.md) — Isolated the native NumPy/JAX likelihood gap to mathematically non-unique PCA axes for a near-isotropic border…
 - [border-relocator-degenerate-pca](2026/08/border-relocator-degenerate-pca.md) — Stabilized near-isotropic border PCA axes with a deterministic axis-aligned branch at a scale-aware eigenvalue…
 - [brain-readability-pass](2026/08/brain-readability-pass.md) — auto-closed on merge

dashboard.html

@@ -204,7 +204,7 @@
 </head>
 <body>
 <header class="hero"><span class="orb"><svg class="mark" viewBox="0 0 48 48" fill="none" stroke="currentColor" stroke-width="1.3" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="24" cy="24" r="20.4"/><path d="M20.2,37.6 L20.2,33.6 C20.2,32.4 19.4,31.6 18.4,30.8 C16.8,29.6 15.8,28.2 15.6,26.6 C15.5,25.6 14.9,25.2 13.9,24.9 C12.6,24.5 12.3,23.7 13.0,22.7 C13.8,21.6 14.5,20.6 14.6,19.4 C15.0,15.4 18.2,12.2 22.4,11.8 C27.2,11.3 31.6,14.6 32.3,19.2 C32.8,22.4 31.6,25.0 30.6,27.0 C29.9,28.4 29.6,29.6 29.6,31.2 L29.6,37.6"/><circle cx="24.6" cy="19.6" r="4.4"/><path d="M22.5,19.7 L24.0,21.3 L26.8,18.1"/><path d="M34.2,16.5 L38.8,16.5 M35.2,21.0 L39.4,21.0 M33.6,25.5 L37.2,25.5"/><g fill="currentColor" stroke="none"><circle cx="40.1" cy="16.5" r="1.4"/><circle cx="40.7" cy="21.0" r="1.4"/><circle cx="38.5" cy="25.5" r="1.4"/></g></svg></span><h1>PyAuto<b>Mind</b><span class="kind">Dashboard</span></h1><div class="rule"></div><p class="tag">Intent. Priority. Flow.</p></header><p class="lede">Every task the Mind is holding. Tap a task's 📋 and its <code>/start_dev</code> command is on your clipboard — paste it into a Claude Code chat to route Claude straight to that task. <a href="#recent">Recent</a> is the same work by date — what has been happening rather than what to do next.</p>
-<ul class="stats"><li><b>3</b><span>In flight</span></li><li><b>3</b><span>Parked</span></li><li><b>5</b><span>Planned</span></li><li><b>134</b><span>Backlog</span></li></ul>
+<ul class="stats"><li><b>2</b><span>In flight</span></li><li><b>3</b><span>Parked</span></li><li><b>5</b><span>Planned</span></li><li><b>134</b><span>Backlog</span></li></ul>
 <div class="fresh"><p><b>Last updated 2026-08-27.</b> This page is generated from <code>active/</code>, <code>draft/</code> and the registry files, so it is only as current as they are. <code>dashboard_refresh.yml</code> re-renders it on every push to <code>main</code> — that heals a stale page, but not a stale prompt: a task that shipped without its prompt advancing to <code>complete/</code> keeps rendering here as pickable backlog. Reconciling those is the refresh below.</p>
 <div class="task"><button class="copy" data-cmd="Bring the PyAutoMind dashboard up to date. Work in the PyAutoMind checkout:
 
@@ -252,7 +252,6 @@ <h3>Quick wins <span class="facets">(small enough, and safe enough to run unatte
 <h2>In flight <a class="mdsrc" href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/active.md">markdown version</a></h2>
 <p class="muted">Issued — each has an open GitHub issue and usually a branch.</p>
 <div class="task"><button class="copy" data-cmd="/start_dev active/16_transformed_message_factor_gradient_unpack.md" aria-label="Copy the Claude command">📋</button><p><a href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/active/16_transformed_message_factor_gradient_unpack.md"><code>@PyAutoFit</code> <code>TransformedMessage.factor_gradient</code> crashes on first call</a> — <a href="https://github.com/PyAutoLabs/PyAutoFit/issues/1501">issue #1501</a><span class="facets"> — issued 2026-08-19</span> — <span class="facets">HOLD — do not start dev. Fix-or-delete hangs off the PyAutoFit#1498 logpdf-contract</span></p></div>
-<div class="task"><button class="copy" data-cmd="/start_dev active/board_without_gh_phase1_seam.md" aria-label="Copy the Claude command">📋</button><p><a href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/active/board_without_gh_phase1_seam.md">Board phase 1: the injection seam, proven on the overnight legs</a> — <a href="https://github.com/PyAutoLabs/PyAutoBrain/issues/303">issue #303</a><span class="facets"> — issued 2026-08-27</span> — <span class="facets">library-dev — not started; routed by /start_dev, no worktree claimed yet</span></p></div>
 <div class="task"><button class="copy" data-cmd="/start_dev active/xla_cpu_eigen_pool_deadlock.md" aria-label="Copy the Claude command">📋</button><p><a href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/active/xla_cpu_eigen_pool_deadlock.md">Why does XLA CPU's Eigen thread pool wedge on the multi_dataset vmap…</a> — <a href="https://github.com/PyAutoLabs/PyAutoFit/issues/1530">issue #1530</a><span class="facets"> — issued 2026-08-27</span> — <span class="facets">not started — research follow-up to the shipped jax-compile-stall epic</span></p></div>
 <h2>Parked <a class="mdsrc" href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/parked.md">markdown version</a></h2>
 <details>
@@ -467,12 +466,6 @@ <h2>Backlog <a class="mdsrc" href="https://github.com/PyAutoLabs/PyAutoMind/tree
 <td class="pick"><button class="copy" data-cmd="/start_dev draft/feature/pyautobrain/board_without_gh_phase2_legs.md" aria-label="Copy the Claude command">📋</button></td>
 </tr>
 <tr>
-<td class="when">2026-08-27</td>
-<td class="what">issued</td>
-<td><a href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/active/board_without_gh_phase1_seam.md">Board phase 1: the injection seam, proven on the overnight legs</a></td>
-<td class="pick"><button class="copy" data-cmd="/start_dev active/board_without_gh_phase1_seam.md" aria-label="Copy the Claude command">📋</button></td>
-</tr>
-<tr hidden>
 <td class="when">2026-08-26</td>
 <td class="what">filed</td>
 <td><a href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/bug/workspaces/gradient_pixelization_pin_residual_drift.md">jax_profiling/gradient/imaging/pixelization.py: 3.2% of its pin move…</a></td>
@@ -712,6 +705,12 @@ <h2>Backlog <a class="mdsrc" href="https://github.com/PyAutoLabs/PyAutoMind/tree
 <td><a href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/maintenance/autolens_workspace_developer/stale_api_rot_audit.md">autolens_workspace_developer: broad stale-API rot (56 symbols, no CI)</a></td>
 <td class="pick"><button class="copy" data-cmd="/start_dev draft/maintenance/autolens_workspace_developer/stale_api_rot_audit.md" aria-label="Copy the Claude command">📋</button></td>
 </tr>
+<tr hidden>
+<td class="when">2026-08-04</td>
+<td class="what">filed</td>
+<td><a href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/triage/nightly_release_blocked_eight_nights.md">Nightly release has been blocked 8 nights running — triage the streak</a></td>
+<td class="pick"><button class="copy" data-cmd="/start_dev draft/triage/nightly_release_blocked_eight_nights.md" aria-label="Copy the Claude command">📋</button></td>
+</tr>
 </table>
 <button class="more" data-page="10">… 10 more (40 left)</button>
 <h2>Epics <a class="mdsrc" href="https://github.com/PyAutoLabs/PyAutoMind/blob/main/epics.md">markdown version</a></h2>