Files
vardkey/ai-note.md
T

59 KiB
Raw Blame History

AI Agent Note — VardKey

Short working handover for the next model. Keep it short.

  • TODO.md (Norwegian) is the requirement specification. Sections 1–18 are the binding spec; 19–31 are history; section 32 is the round in progress. The full reasoning is in the git history.
  • The binding rules live in TODO.md and README.md. This note does not repeat them — it carries the handover, the rules that are easy to get wrong, and the traps that already cost time.
  • HANDOVER.md was deleted as outdated. Do not recreate it.

Status

  • Done and verified: sections 1–31 and section 18 steps 1–11. Section 32.1, the icon, is done — the SVG drawing, the favicon on every page, the top-bar brand mark and the two 192 and 512 px PNGs — together with section 32.2, the UI pass with the retro theme, section 32.3, the alert and page texts, and section 32.4, the bug report (round 32-A and round 32-B). The follow-up round of 32.1 and 32.2 — the favicon, the top-bar brand mark, the footer at the bottom, the single way back, the history button and the three line/table changes in the history view — is done too. Schema version 12. 498 green tests. The StartOS package is built, sideloaded and running on a StartOS 0.4.0 box; the package version is still 0.1.0:0.
  • What is left of section 32: 32.6 the documentation drift, and 32.5 the raise to 1.0.0:0 together with section 18 step 12. 32.6 has to come after 32.1–32.2 and before 32.5. One item inside 32.4 is left to the user: BR-003 — read the 32.4 section below before acting on it, because the report's own wording for it is out of date. The user's own manual pass is also outstanding: the retro theme on a narrow phone, the footer on a short page, the favicon in the tab, the brand mark in the top bar, installing the PWA with the two PNG icons, the history view after the three changes, and the whole box test again.
  • A round stops when it is done. Do not start the next item unless the user asks.
  • Verified how: the automated suite, the user's manual browser testing, and the install and test round on a StartOS 0.4.0 box. The user runs the suite; the agent has no shell.
  • Tests: python3 -m unittest discover -s tests. They run on a Windows copy at C:\Users\<brukarnamn>\Downloads\helsesjekk, so they must pass there as well as on Linux.
  • Known gap, still open, folded into 32.6: docs/application.md and README.md still describe the API as it was before section 31 — no login/session endpoints, no gating or CSRF rule, the old maintenance command name, the undated export name, and now also the two new summary fields, the moved field reminder, the added model field in the check detail, the icon link on the pages and the two PNG icons. The documentation conflict in the bug report (BR-004) covers the same ground.

Section 32 — what is left

  • 32.1 The icon — done. icon.svg at the root and static/icon.svg are the same retro drawing: a raised grey window with the three keys, centred in the face. The navy title bar is gone, and the keys were recomputed so they are centred: one key group is 21 high, the face inside the bevel runs from 6 to 122, so the centre is 64 and the keys sit at translate(23, 25.5), (23, 51.5) and (23, 77.5). The root file is 1.2 KiB, well under the 40 KiB cap, and it is plain SVG with no font, image or network reference, so the default-src 'self' CSP holds. Every one of the six pages links /icon.svg as its favicon, the top-bar brand mark is the icon image under the same brand-mark class, and the two PNGs the user rasterized — static/icon-192.png and static/icon-512.png — are listed in the manifest after the SVG, served through STATIC_FILES and cached by the service worker. /icon.svg was already in STATIC_FILES, in the manifest's icons entry and in the service worker's allow-list, so no path is new; the two PNG paths are.
  • 32.2 UI polish — done, including the follow-up round. See the round 32.2 section below. The only thing left for the user is the manual browser pass on a narrow phone and over the new points.
  • 32.3 Alert and page texts — done. See the round 32.3 section below.
  • 32.4 The bug report — done. See the section below for what each finding turned out to be.
  • 32.6 The documentation drift. README.md, docs/application.md, instructions.md, UPDATING.md, AGENTS.md and the package's own i18n strings are all behind the code. TODO.md §32.6 lists what is wrong in each. It answers the rest of BR-004, and it must be done after 32.1–32.2 and before 32.5, so the documents describe the released state.
  • 32.5 The version raise. 0.1.0:0 → 1.0.0:0, once the items above are done and the user says so. No schema change, no new endpoint, no new feature. startos/versions/ stays empty, because the app migrates its own schema.

Round 32.2 — the UI pass and the retro theme

static/styles.css was rewritten as a Windows application of the mid-1990s: background #c0c0c0, fields #ffffff, dark edge #808080, darkest #000000, a navy #000080 title bar with white text, and no rounded corner, shadow or gradient anywhere. Depth comes from two-tone bevels only. Headings are title bars, buttons are raised and go sunken on :active, fields are sunken wells, focus is the classic outline: 1px dotted #000000, and the tables are a 1 px grid. The responsive rules and every selector the scripts and the tests rely on are unchanged, including the exact string .rotation-list { margin-top: 22px; }.

  • The only border-radius in the sheet is an explicit reset to 0. It is there so a browser's own rounded control never shows through, and the test asserts exactly that: every border-radius value found in the sheet must be 0. Do not drop the reset, and do not add a rounded corner.
  • theme-color and the manifest move together. The five pages, /login and theme_color/background_color in static/manifest.webmanifest are all #c0c0c0 now, and tests/test_pwa.py asserts that.
  • The key card changed shape. The number and the type share the top line (01 · Hardware via keyTypeLabel()), the type has no line of its own, the path stands alone in italics (.key-path) with no Address path: prefix, the model is a line of its own under the name with no Model:/Service: prefix and is left out entirely when it only repeats the key's name (sameLabel() normalizes case and whitespace), and only Next check due is shown — Last checked is gone. tests/test_key_status.py and tests/test_health_check_form.py pin the new order, so a fourth line or a reordered card breaks them.
  • The dashboard heading is dynamic. p#setup-shape in index.html carries the fallback Bitcoin Health Monitor for a Multisig Setup., and renderSetupHeading() in app.js replaces it with Bitcoin Health Monitor for a ${threshold}-of-${key_count} Multisig Setup. from /api/setup. The static description and the whole section.setup-card are gone, and setupTitle, setupDescription and renderSetupShape() were removed with them, so no null reference is left. The manifest description stays Bitcoin Multisig Health Monitor (section 29) — only the dashboard line is dynamic. tests/test_app.py asserts that the dashboard no longer contains the static description.
  • The field reminder moved. Never paste a seed phrase or a private key here. now stands once, as a p.secret-warning above the whole key list in setup.html. It is gone from health-check.html (3 static fields), health-check.js (3), settings.js (2) and setup.js (one per card). secretWarning() no longer exists in setup.js at all. The server is untouched: secret_guard.py, validation.py and auth.py still refuse a secret in every free-text field. WarningTests in tests/test_secret_guard.py was rewritten to pin the new rule (5 tests, down from 6), and TODO.md §2, §12 and §15 were reworded.
  • Two review lists went. Needs assessment is gone from the Review summary in health-check.js (the form itself shows what is missing), and so is the Key N: … · Check date: … · Next check due: … line. Results requiring follow-up stays, and tests/test_health_check_form.py now asserts assertNotIn("Needs assessment", script).
  • The history list card changed. It shows the date, Key N: <name> for each key (from the new key_names), and one short status, All OK or Needs follow-up (from the new needs_follow_up). Next check due and the Software text are gone from the card; the read-only detail view keeps every field it had.
  • The two new summary fields are additive. GET /api/health-checks still returns everything list_health_checks() returns and adds key_names (the name each key has now, in the same order as key_numbers) and needs_follow_up (whether that check raised anything to follow up, from build_alerts() — the same rules the dashboard uses). needs_follow_up is a property of the saved check itself, so closing an alert or renaming the key later does not change it; that is deliberate and pinned by test_the_history_list_marks_a_check_that_needs_follow_up. The enrichment lives in app.py's _send_health_check_list, so database.py is untouched.
  • The tables stay inside the page. table-layout: fixed and overflow-wrap: break-word on .detail-table, plus a smaller font and tighter padding in the max-width: 700px query, keep Security status and Follow-up closed from running off a narrow screen. CSS only, no new JavaScript.
  • Local service and Your data stays local. are gone from the top bar and the footer on all six pages, including /login.
  • Tests: tests/test_health_check_form.py (+6), tests/test_app.py (+2, one renamed), tests/test_secret_guard.py (WarningTests rewritten, 6 → 5), tests/test_key_status.py (one test renamed and extended) and tests/test_pwa.py (colours). The suite went from 477 to 484.
  • The help-line test is page-aware on purpose. settings.html has no class="field-grid" at all, so the test maps each of the three pages to the block it actually puts under the help line: a field grid on the setup page, a stack on the check form and on the settings page. Do not "simplify" it back to one block name for all three.

Seven small changes, no new feature and no schema change. The reasoning is in TODO.md §32.1 and §32.2 points 16–21.

  • The icon lost its navy title bar and the keys are centred. icon.svg and static/icon.svg are the same drawing. The keys are 21 high per group and the face is 6–122, so the centre is 64 and the translate values are 25.5, 51.5 and 77.5. The style, the viewBox and the role/aria-label are unchanged.
  • The two SVG files must keep the same drawing. The package icon at the repository root is the source of truth, and static/icon.svg is the copy the PWA serves. Only the leading comment differs, and test_the_served_icon_carries_the_same_drawing_as_the_package_icon compares them from the <svg> element onwards. A change to one drawing has to be made in both.
  • The two PNGs are rasterized from static/icon.svg and are real files in the repository. static/icon-192.png and static/icon-512.png were made by the user, because the agent has no rasterizer. IconTests reads each file's IHDR and asserts 192×192 and 512×512, so a PNG made at the wrong size fails the suite instead of silently breaking the install prompt.
  • The favicon is on all six pages. <link rel="icon" type="image/svg+xml" href="/icon.svg"> sits in <head> next to the manifest link on the five pages that have one; /login has no manifest, so it stands next to the stylesheet there.
  • The brand mark is the icon image. <span class="brand-mark" aria-hidden="true">V</span> became <img class="brand-mark" src="/icon.svg" alt="" aria-hidden="true">. The class name is kept, because the stylesheet and the tests refer to it. .brand-mark is now width: 18px; height: 18px; border: 2px solid; border-color: #ffffff #000000 #000000 #ffffff; background: var(--face); image-rendering: pixelated; — display: grid, place-items: center, color, font-size and font-weight are gone, and IconTests asserts that place-items: center appears nowhere in the sheet.
  • The footer sits at the bottom on a short page. body is a column with min-height: 100vh and min-height: 100dvh, and main { flex: 1 1 auto; } takes the rest, so .footer lands at the bottom. No other layout rule was touched. IconTests pins those four declarations.
  • There is one way back to the dashboard. The top-bar Back to dashboard link is gone from history.html, settings.html, setup.html and health-check.html; the VardKey brand mark is the way back. The link in the blocked-keys note stays — it points at the dashboard for a different reason, and tests/test_health_check_form.py still asserts its wording. history.html also lost the Start annual health check button: a check is started from the dashboard.
  • The history view changed in three places. In static/history.js, mergedLine(...) joins wallet_software and software_version into one line (Sparrow Wallet 2.5.1) and model and firmware_version into one line (Blockstream Jade 1.4.1); a missing part is left out and a pair with neither shows —. The Follow-up closed table is now ["Alert", "Action", "Closed"] — the Source column is gone.
  • The model had to be added to the detail response. get_health_check() did not select k.model, so GET /api/health-checks/<id> carried no model at all. It is now selected in the key-check query and comes in in addition to the fields that were already there; no existing field was renamed.
  • Tests: three in tests/test_alerts.py, four in tests/test_pwa.py and three in tests/test_health_check_form.py in the first half, plus four more in tests/test_pwa.py when the PNGs arrived — so the suite went from 484 to 498. Nothing in tests/test_app.py or tests/test_database.py had to change: the added field is additive and no test there asserts a whole record.

Section 32.4 — the bug report, as it turned out

BUG-RAPPORT-AI-AGENT.md carries the full verdicts at the end. In short:

  • BR-001 — false positive. The source files are in main and tracked by Git. The report was written from the repository's web view, not from a checkout.
  • BR-002 — confirmed and fixed (round 32-B). See the guard section below. The user has since confirmed by hand that a seed phrase is stopped in the name field and in the address path field.
  • BR-003 — the report's wording is out of date; what is left of it belongs to §32. The report says the manual end-to-end pass is still open, but it quotes a §31 text that no longer exists, and §16 and §17 record that the manual browser pass and the box tests were run and verified through §31. What is actually outstanding is the new manual work §32 adds — the icon (32.1), the UI pass and the retro theme (32.2), the alert texts (32.3), the follow-up round's pages, and the whole box test again after the changes (32.5) — plus the user's confirmation that they have been run. Do not re-verify what §16 and §17 already record as done.
  • BR-004 — falls away. AI-STATUS.md and DATA-MODEL.md do not exist in this repository, so there is no conflicting pair to reconcile. The real documentation drift is 32.6.
  • BR-005 — verified in code, no behaviour change. Only GET /health, GET /api/session and /login are open; every other /api/* answers 401 and the pages answer 302 to /login; every state-changing POST needs a session and the CSRF token (403). The Secure cookie attribute is deliberately left off, because StartOS terminates TLS and proxies plain HTTP into the container. The report itself says not to change app.py behaviour for this one.
  • BR-006 — confirmed and fixed. The CSV export neutralizes a text cell a spreadsheet would read as a formula.
  • BR-007 — confirmed and fixed. The service worker only stores a successful, same-origin, non-redirected response.
  • BR-008 — confirmed and fixed. The address-path comparison is canonical.
  • The report was written against a different file set. It asks for SECURITY.md, PRIVACY.md, THREATMODEL.md, PLATFORM-ACCESS.md, ENCRYPTION.md, OFFLINE.md, AI-STATUS.md and DATA-MODEL.md; none of them exist here. Do not create them.
  • The counts in the report were partly wrong and are corrected there. Round 32-A added ten tests, not eight, and tests/test_secret_guard.py holds 43, not 44: 5 wordlist tests, 10 detection tests, 17 field-level tests, 6 warning tests and 5 API tests. The suite went from 422 to 475, which is the number the user confirmed. Round 32.3 added two more — one in tests/test_alerts.py and one in tests/test_health_check_form.py — and round 32.2 added seven net, for 484. The seven are six new tests in tests/test_health_check_form.py and two in tests/test_app.py, minus the one WarningTests test that the rewritten class no longer has. The follow-up round of 32.1 and 32.2 added fourteen more in total, for 498.

Round 32.3 — the alert texts and the page texts

The eleven rows in TODO.md §7 (eight codes, two of which have a key form and a software form) now read alike in shape: what is wrong, which key or software it is about, and what closes the alert. code, severity, source, the actions and the software field are unchanged. The shape is written down in §7, so a new alert follows it.

  • The source names the key or the software, so the message does not repeat it. source is either Key N: <name> or the software name, and it is shown right above the message in the alert list, in brackets on the key card and in front of the message in the form's blocking note. That is why the messages say "this key". The one exception is the cloud-key form of software_incompatible: its source is the key, so its message names the software. The follow-up round made the security alerts follow this rule too — see the alerts section below.
  • The two forms of a security code carry their own message. _VULNERABLE_MESSAGE and _COMPROMISED_MESSAGE were split into _VULNERABLE_KEY_MESSAGE / _VULNERABLE_SOFTWARE_MESSAGE and _COMPROMISED_KEY_MESSAGE / _COMPROMISED_SOFTWARE_MESSAGE, because the key form is closed by a rotation and the software form by removing the software.
  • Two more constants were pulled out of _key_alerts: _HARDWARE_FAILURE_MESSAGE and _SEED_BACKUP_MESSAGE. _SOFTWARE_INCOMPATIBLE_KEY_MESSAGE is a template with {software}.
  • "Replace the key" is gone where the only action offered is rotate. "PROTOCOL 3: key loss" and "follow the key loss procedure" are gone too — neither term is defined anywhere in the app or the documents, so they pointed at a concept the user does not have.
  • The contact alert no longer names the key in its message. It says "this key"; the source carries the name.
  • The remaining duplication is deliberate: the action appears in the message, on the button in the alert list and in the form's Close it with: … line. The button and the note say what the user can press; the message has to stand alone when it is read without them.
  • The blocking 409 message no longer prints the internal code. It reads "This key has an open alert. Close the alert before recording a new check."
  • The page texts that went with it: the duplicated Multisig setup eyebrow in index.html, the duplicated Settings eyebrow in settings.html, the duplicated Sign in heading in login.html, the blocked-key help line in health-check.html that repeated the note above it, the static #software-help text that the script overwrote before it was ever read, the two Removing an entry… helps in settings.html that became one paragraph, the Open settings wording in setup.js that is now Open the settings like the dashboard's, and the repeated privacy sentence in app.js's setupDescription. validation.py's "a configured cloud key already provides this wallet software name" is now "this name already comes from a configured cloud key".
  • Not merged on purpose: ACTION_LABELS / ACTION_CONFIRMATIONS stay in three copies (static/app.js, static/health-check.js, static/history.js). They are identical today; a shared file would have to be added to STATIC_FILES in app.py and to the allow-list in static/sw.js.
  • UNAVAILABLE_MESSAGE in health-check.js stays, even though the section it is rendered into is hidden for a key that could not be checked. It is the defensive render, and a test pins it.

Round 32-B — the wallet-secret guard

secret_guard.py is the one place that decides whether a piece of text looks like a wallet secret. It is local (no network call), silent (the text never reaches a log, a message, or an exception), and narrow (it matches only the shapes that are unmistakable, so an ordinary note is not refused).

  • What it recognises: a PEM private-key block; an extended private key with any SLIP-132 prefix (xprv, yprv, zprv, tprv, uprv, vprv) at 100 base58 characters or more; a WIF key (51 or 52 base58 characters starting 5, 9, K, L or c); a BIP38 key (6P, 58 characters); and a run of 12 or more consecutive BIP39 words, also in the middle of a sentence.
  • What it returns: a short English label such as "a seed phrase", never the value. rejection_message(field, label) builds the sentence the user sees, and it never repeats the value either.
  • The wordlist is embedded, 2048 words, from bitcoin/bips bip-0039/english.txt. It is data, not a dependency, and the app still makes no network call. tests/test_secret_guard.py checks the list: 2048 words, alphabetical, 3–8 lower-case letters, and no two words sharing their first four letters. That last check is the strong one — the specification guarantees it, so a mistyped or missing word breaks it. Do not edit the list without that test.
  • Where it is wired in: validation.py calls _reject_secret() from the four text helpers _text, _key_name, _model and _required_path, so every free-text field is covered by construction — the key name, the model, the wallet software name, the software version, the firmware version, the four comment fields, and the address path. auth.py checks the admin username the same way.
  • The refusal is a ValidationError. SecretDetected subclasses it, so the existing handlers answer 400. Had it been a plain ValueError, STORAGE_ERRORS would have caught it and answered 500.
  • app.py needed no change at all. Do not add a handler for this.
  • The password is deliberately not checked. A long passphrase is a legitimate password, and refusing one would lock the user out of their own account.
  • The guard runs before the field's length limit. A seed phrase pasted into a 60-character label field is answered with "looks like a seed phrase", not with "too long" — the other order would invite the user to shorten a seed phrase instead of removing it. The limit is still the outer bound and still applies to everything the guard lets through. This order is deliberate and pinned by tests; do not swap it back.
  • The reminder in the UI changed in 32.2. Round 32-B put Never paste a seed phrase or a private key here. next to every free-text field, with the class .secret-warning in static/styles.css. It now stands once, above the whole key list in static/setup.html, and nowhere else; the per-card and per-field copies are gone. The server side is unchanged. The page notice is not the same thing and stays on health-check.html, setup.html and settings.html. tests/test_secret_guard.py's WarningTests pins the new rule: one class="secret-warning" in setup.html, none in the check form or the settings, and no secretWarning in setup.js.
  • The limitations are documented in the module docstring and are not to be played down: a bare 64-character hex string is not refused (it is indistinguishable from a transaction id, which is legitimate note content); only the English wordlist is used; a secret split across several fields, or in a shape outside the patterns, is not caught; and a false positive is possible for a note that happens to be twelve or more BIP39 words in a row.

Section 31 — the points before version 1 (done)

Seven points, done in the order 7, 5, 4, 3, 2, 1, 6 — the data model first, because everything else builds on it. The icon was split out into section 32.

  • Schema 12. app_settings (multisig_threshold, multisig_key_count, admin_username, admin_password_hash, session_secret) and a nullable keys.model. The key_number CHECK was loosened through a table rebuild with PRAGMA foreign_keys = OFF and PRAGMA foreign_key_check before commit.
  • The admin login. auth.py is the only place that knows the credential: hashlib.scrypt with a random salt per password, a signed session cookie (HMAC-SHA256 over an expiry, keyed by session_secret, no session table), and a double-submit CSRF token. Every route except GET /health needs a session: 401 on the API, a redirect to /login on the pages, and a CSRF requirement on every state-changing POST.
  • maintenance.py, with delete-all-data and set-admin-password.
  • The installable PWA, without offline data: static/sw.js (an allow-list of shell paths only), static/manifest.webmanifest, static/icon.svg and the two PNGs beside it, the registration and offline state in static/app.js, and the manifest link on the five pages.
  • The flexible threshold (2 ≤ key_count ≤ 15, 1 ≤ threshold ≤ key_count), still for one setup only.
  • The dated export name, YYYY-MM-DD-vardkey-export.csv. The download attribute in static/index.html was removed, because one with a value overrides the server's Content-Disposition filename.

Section 30 — the StartOS package

The package is one service: one daemon, one container, one SQLite database on the persistent main volume, mounted at /data. No Postgres, no Redis, no external backend, no extra container. The decisions are in TODO.md section 30.

  • The application and the package share one repository, at the root. That is a deliberate deviation from the guide's "a package repo is not a fork of the application": the rule assumes a published upstream image or a git submodule, and VardKey has neither. AGENTS.md states the deviation so nobody adds an upstream-project/ submodule to "fix" it.
  • Anything the Dockerfile does not COPY is not in the image. The modules go in with COPY *.py ./ and the assets with COPY static ./static; .dockerignore keeps tests, data, node_modules, javascript and the build output out of the context. static/ must be copied: app.py answers /health from code, but every page is read from ROOT/static, so an image without it reports a healthy service whose whole UI answers 404.
  • Three things must agree. startos/utils.ts holds uiPort (8080); startos/main.ts passes it as PORT and uses it in the readiness URL; startos/interfaces.ts binds it. Change one alone and the readiness check never goes green. The DATABASE_PATH constant in the same file is what puts the database on the volume. The readiness check is checkWebUrl against http://127.0.0.1:8080/health — and /health must stay open, or StartOS reports the service as unhealthy. A red ECONNREFUSED 127.0.0.1:8080 blip before VardKey listening on port 8080 is just the first poll landing before Python has bound the socket; nothing to fix.
  • Two actions, registered through Actions.of() in startos/actions/index.ts, with their strings in startos/i18n/dictionaries/: deleteAllData.ts (delete-all-data) and setAdminPassword.ts (set-admin-password). Both run maintenance.py in a withTemp subcontainer, behind a confirmation the handler enforces. The generated password never travels into the container: maintenance.py generates it there and prints it, which is the packaging guide's own password-reset pattern.
  • Backups. startos/backups.ts is sdk.Backups.ofVolumes('main'). StartOS stops the service for the duration of a backup, so the database file is quiescent while it is copied — the open question section 30 raised needed no snapshot hook, and it is not a known limitation.
  • No CI. The guide assumes GitHub Actions; this repository is on Gitea, so there are no .github/workflows/ and there must not be any. The .s9pk is built locally with make.
  • The skeleton (package.json, tsconfig.json, Makefile, .prettierrc, startos/sdk.ts, startos/index.ts, startos/i18n/, startos/fileModels/.gitkeep, assets/.gitkeep) was written to match the Start9 template rather than generated by s9pk init-workspace, because that command was not available where it was written. The successful build and install confirm they are equivalent; @start9labs/start-sdk is pinned to 3.0.4.
  • The root icon.svg is the package artwork. Nothing in startos/ names it; the build reads it from the repository root. It is the same retro drawing as static/icon.svg, and it must stay under 40 KiB.
  • Documentation layout. README.md is the StartOS package README, with the guide's fixed heading set. The application's own reference moved to docs/application.md, which instructions.md links to under ## Documentation. AGENTS.md is the repo-specific agent context, and CLAUDE.md is the one-line @AGENTS.md import. Both README.md and docs/application.md are behind the code — see the gap under Status.

Alerts, closures and key identity — the parts that are easy to get wrong

The rules are in TODO.md §7; these are the traps that have already cost time.

  • The acknowledgement stores the key. alert_acknowledgements.key_id is nullable and holds the key the alert belonged to, or NULL for a software alert raised by the settings list. An alert with a key is acknowledged and hidden by (code, key_id); an alert without one stays on (code, source). This closed a real bug: a rename — or a rotate with keep_name=false — used to reopen a closed alert, because source carries the name the key has now.
  • A key-sourced security alert carries Key N: <name>. build_alerts() reads the key checks into two lookups — key_ids (name → id) and key_sources (name → Key N: <name>, built with _key_source()) — and _security_alerts() takes both. When the provider is the key's own configured name, the key is the source and the alert belongs to it, so it is closed by a rotation; otherwise the source is the plain software name and it is closed by removing that software from the settings. The software form's source is unchanged. Because the acknowledgement hides a key alert by key_id, not by source, an alert closed before this change stays closed — test_a_key_named_security_alert_is_hidden_by_its_key pins it. rotation_alert_targets(), software_alert_targets() and newest_alert_check_id() all read the same build_alerts(), so they see the same source, which is what keeps an alert closable.
  • The contact alert belongs to the checked key. It gets key_id and source = Key N: <name>, and its message says "this key" and no longer repeats the name (round 32.3). The severity stays info and the action stays contact_updated. It blocks its key like every other key alert, because blocking_alert() and the blockingAlerts filter in the form only look at key_id. A record without a key check raises no contact alert at all.
  • software_incompatible has two forms. A software from the settings list keeps source = the software name, key_id = None, actions [remove_software], and carries software. A wallet software that is the name of a configured cloud key can never be removed from the settings, so that form belongs to the checked key instead: key_id = the checked key, source = Key N: <name>, actions [replace_hardware], and no software value, so a settings removal can never close it. Only that form names the software in its message, because its source is the key.
  • alerts.py needs the cloud key names. build_alerts() and collect_alerts() take cloud_names=(), threaded through rotation_alert_targets(), software_alert_targets(), and the four call sites in app.py (_send_alerts, _blocking_alert, _rotate_key, _close_removed_software), plus newest_alert_check_id(). app.cloud_key_names(keys) builds the tuple. If newest_alert_check_id() does not get exactly the same names as collect_alerts(), the alert's source differs, the acknowledgement stores check_id = None, _is_acknowledged() refuses to hide an alert without a check id, and the alert can never be closed. Keep the default () on build_alerts(). _send_health_check_list() passes the same tuple for its needs_follow_up field, for the same reason.
  • The closures that belong to a key are shown. app.closed_alerts() and database._grouped_closures() both filter on key_id IS NOT NULL, so a rotation, a replaced hardware, an updated contact, and an ignored key alert are listed, while remove_software and a software-sourced ignore are not. GET /api/health-checks/<id> returns them as closed_alerts, and the CSV has one last column, Follow-up closed, with code: action (YYYY-MM-DD), hanging on the check it closed through check_id. The history table shows Alert, Action and Closed only — the source is not a column.
  • The key card shows the next due date with the alert note. createKeyCard() in static/app.js always renders Next check due and adds the alert note on top. Last checked was dropped in 32.2, but the note still comes in addition to the date, never instead of it — that is what cost the dates once before, for a key whose only alert was the informational contact one.
  • The type is part of a key's identity. The setup locks it, the settings never change it, and only a rotation may pick the other one: POST /api/keys/<id>/rotate takes an optional type, and leaving it out keeps the current type. save_key_settings(..., key_type=None) uses COALESCE(?, type), so the settings and a rotation share one statement.
  • A key that becomes a cloud key must not take a taken name. A cloud key offers its own name as a wallet software choice, and the settings list holds the other choices, so a name the list already holds is refused with 409 through WalletSoftwareNameInUse — a ValidationError subclass, so it has to be caught before the general ValidationError in _rotate_key, or the 409 turns into a 400. validate_rotation() therefore takes the wallet software list as a fourth argument.
  • A rename needs no rotation and no confirmation. In the rotation section the name can be changed on its own through POST /api/settings/keys, for any key, and the address path of a key with no saved check can be changed there the same way. Marking a key with Rotate this key and ticking the confirmation box sends it through the rotate endpoint instead, together with the type and the new path. The single button is Save keys.

Deliberately left out

remove_software is still a dead end in two places: a software-sourced security_vulnerable offers it next to ignore (so the alert stays closable, but one button leads nowhere), and a software-sourced security_compromised offers it as its only action, so an alert whose "software" is a cloud key name can never be closed. Neither is fixed.

Rules that are easy to get wrong — do not re-open

  • One submission records one key. There is no "round" grouping, and each key's next check is its own date plus one calendar year, computed on the server and not editable in the form.
  • A key that could not be checked records nothing else. That covers available = 0 and a recorded address path that does not match. wallet_software and ai_security_checks must be empty; a non-empty value is a 400, not something dropped silently. Nothing is stored for them, and software_compatibility_status becomes NOT_ASSESSED.
  • Availability is required and offers only Yes/No. A cloud key collects only Availability; its compatibility is derived from it, and its own name is the wallet software.
  • A hardware key needs an explicit compatibility status, and Hardware works needs an explicit Yes or No. A missing value, "" and NOT_ASSESSED are all a 400.
  • The contact confirmation is required for every submission, including one for a key that could not be checked, and it has no "not assessed" value.
  • NOT_ASSESSED is not offered or accepted anywhere else. No field but the derived software_compatibility_status accepts it.
  • Address path verified is never a form field again. The server derives it from the comparison.
  • The address path is locked after the first saved check, and no two keys may share one — in the setup, in the settings or in a rotation. The comparison is canonical: m/48h/0h/0h/2h and m/48'/0'/0'/2' are the same path, and the stored spelling is never rewritten.
  • The type is locked by the setup, and only a rotation may change it. The settings never take a type, and a rename alone is not a rotation.
  • An alert is closed by the user, not by a newer check. A newer check that raises nothing does not clear an older alert, and a key with an open alert cannot get a new check at all.
  • An acknowledgement stores the action name, not a free reason, and only the actions the endpoint performs are accepted. rotate has its own endpoint, because it requires a new address path.
  • The export is plain CSV only. XLSX is not built and the app does not promise Excel compatibility. The column names and the column order are an agreement; do not change them. A text cell that a spreadsheet would read as a formula is neutralized with a leading apostrophe, and that is the only change a cell may receive.
  • The PWA cache stores only a valid shell answer. isCacheable() in static/sw.js requires response.ok, response.type === "basic" and !response.redirected, so a login page (or an error page) can never replace the cached shell. The allow-list is the other half: the worker never names /api/*, /health or /login.
  • A free-text field refuses a wallet secret, and the guard runs before the length limit. See the 32-B section above. The refusal is a ValidationError, the message never repeats the value, and the password is not checked. The reminder that says so is one line in the key setup, not one per field (32.2).
  • Every alert message says what is wrong, which key or software, and what closes it, and it does not repeat the name in source — except the cloud-key form of software_incompatible. See the round 32.3 section above and TODO.md §7.
  • The two new summary fields are additive and mean what they say. key_names follows key_numbers; needs_follow_up is a property of the saved check, so it does not change when an alert is closed or a key is renamed. Do not "fix" it to follow collect_alerts(), which hides acknowledged alerts and would rewrite history.
  • The prohibition lists in sections 1 and 13 of TODO.md stay unticked. They are standing prohibitions, to be re-confirmed before release — with the one amendment for the installable PWA.
  • Whether a date is overdue is decided in the browser, so /api/key-status stays independent of the current date.
  • /api/previous-values still returns a security map that nothing consumes any more, and its address_path per key is no longer used by the form, which pre-fills the path from keys.address_path. Both are deliberate: removing them would shrink the endpoint, but nothing breaks while they stay.
  • delete_all_data() deletes the checks before the keys. key_checks.key_id is ON DELETE RESTRICT, so the other order fails. The acknowledgements are deleted explicitly, because they have no foreign key at all.

Data model (schema version 12)

  • keys: key_number (1–multisig_key_count), name, type (hardware/cloud), model, address_path, created_at. There is no device column. The name is editable in the settings; the type and the address path are the identity of the key, the path must be unique across the rows, and the type only changes through a rotation. model is optional documentation and is not locked.
  • health_checks: checked_at, next_check_due, wallet_software, software_version, software_compatibility_status, software_comments, contact_info_verified, comments, plus the legacy sparrow_version / nunchuk_version. wallet_software and contact_info_verified are free text.
  • key_checks: one row per submission — available, hardware_works, firmware_version, seed_backup_verified, address_path_verified, address_path, comments. The two cloud columns are gone.
  • ai_security_checks: provider (free text), status, comments, with UNIQUE (health_check_id, provider).
  • wallet_software: name, created_at. Stored whole, seeded with DEFAULT_WALLET_SOFTWARE (Sparrow Wallet) when the list is empty, never emptied through the API. delete_all_data() empties it, and the seed puts the default back.
  • alert_acknowledgements: code, source, key_id, check_id, reason, acknowledged_at. reason holds the action name that closed the alert, and key_id holds the key the alert belonged to, or NULL for a software alert from the settings list.
  • app_settings: name (PRIMARY KEY), value. Holds multisig_threshold, multisig_key_count, admin_username, admin_password_hash and session_secret. delete_all_data() removes the two multisig rows, so the 2-of-3 default applies again, and keeps the three credential rows.
  • tool_checks no longer exists.
  • Startup migration rebuilds keys when it still has device or the old key_number CHECK, key_checks when it uses the old layout or still holds the two cloud columns, and health_checks and key_checks when their CHECK constraints still allow the old values. It then drops tool_checks if it is there, and adds the newer columns through OPTIONAL_COLUMNS.
  • The table definitions are {table} templates, so a rebuild can create the same table under a temporary name. health_checks is a parent table, so the whole migration runs with PRAGMA foreign_keys = OFF, checks PRAGMA foreign_key_check before committing, and rolls back on any failure.
  • tests/test_database.py carries LEGACY_SCHEMA, SCHEMA_VERSION_5, SCHEMA_VERSION_8, SCHEMA_VERSION_9, SCHEMA_VERSION_10 and SCHEMA_VERSION_11 as literal older schemas. Copy that pattern for a new migration test: write the old schema, run initialize_database(), then assert on the migrated rows.
  • There is no data migration for an older test database: the user deletes the data folder and configures the setup again. An acknowledgement written before version 11 has no key, and the older contact alert used the source Contact information, which no longer matches, so the alert it closed is raised again. Section 29 renamed the database file for the same reason, and the old file is discarded with the rest.

API

Pages: /, /setup, /settings, /health-check, /history, /login. Endpoints: GET /health, GET /api/session, POST /api/admin, POST /api/login, POST /api/logout, GET /api/setup, POST /api/setup, GET /api/keys, GET /api/key-status, GET /api/previous-values, GET /api/settings, POST /api/settings/keys, POST /api/settings/wallet-software, POST /api/keys/<id>/rotate, GET /api/health-checks, GET /api/health-checks/<id>, GET /api/alerts, POST /api/alerts/acknowledge, GET /api/export, POST /api/health-checks. Field details are in docs/application.md — which is behind the code, see the gap under Status.

GET /api/health-checks returns each summary with id, checked_at, next_check_due, software_compatibility_status, wallet_software, contact_info_verified, comments, key_numbers, and the two added in 32.2: key_names and needs_follow_up.

Maintenance commands, run inside the container by the StartOS actions and usable by hand: maintenance.py delete-all-data and maintenance.py set-admin-password. Neither is an HTTP endpoint.

The rotate payload is replaced + keep_name + name + type + address_path, and type is optional. GET /api/health-checks/<id> also returns closed_alerts, and each of its key_checks rows carries model as well — the field was added in the follow-up round of 32.1 so the history can show the model and the firmware version on one line, and the fields that were already there keep their names. The settings page has two sections, and POST /api/settings/keys is what a rename alone goes through.

Five things that are easy to break:

  • New static files must be added to STATIC_FILES in app.py, or they return 404. That includes /settings, /settings.js, /login, /login.js, /session.js, /sw.js, /manifest.webmanifest, /icon.svg and the two raster icons /icon-192.png and /icon-512.png.
  • /api/setup, /api/settings, /api/alerts, /api/key-status, /api/previous-values and /api/session are extra endpoints that are not in the minimum list in section 11.
  • STORAGE_ERRORS in app.py includes RuntimeError, because list_keys() raises it for a partial setup. Keep it there, or a broken database answers with a traceback instead of a 500. It also includes ValueError, so validate_admin_credentials() has to be called in its own try and mapped to a 400 before the storage handler — and it is why a refused secret has to be a ValidationError rather than a bare ValueError.
  • The settings, rotation and acknowledgement endpoints keep the same error handling as the existing ones: a bad payload is a 400, an unknown key or alert is a 404, a locked or duplicate address path is a 409, and a broken database is a 500. DuplicateAddressPath and WalletSoftwareNameInUse must be caught before the general ValidationError — in _save_key_settings and _rotate_key respectively — or the 409 turns into a 400. A missing or invalid session is a 401, and a missing CSRF token is a 403.
  • newest_alert_check_id() in app.py derives the check id for the acknowledgement, the rotation and the software removal, so the same comparison rule is used in all three places. blocking_alert() is the module-level helper the 409 is built on, and _blocking_alert() reads the key id straight from the payload, so the rule applies before validation.

Traps that already cost time — do not repeat

  • with sqlite3.connect(...) does not close the connection. It only commits or rolls back the transaction. On Windows the database file stays locked and TemporaryDirectory.cleanup() fails with WinError 32. Always close explicitly with try/finally; see write_schema() in tests/test_database.py.
  • A module the image is missing is a runtime error, not a build error. auth.py was left out of the Dockerfile because the module list was written before it existed, and both app.py and maintenance.py then died with ModuleNotFoundError inside the container while every test stayed green. The COPY *.py ./ line is what stops that recurring. After any change to the image, check the modules, not just the tests. secret_guard.py is covered by the same line.
  • Every server test module needs a session, and a redirect makes a missing one look like a pass. A page fetched without a session answers 302 → /login, and urllib follows it, so an assertion on page markup reads the login page's HTML instead of failing with 401. Use tests/server_session.py's SessionMixin, and note that the manifest and the icons are open while the pages are not.
  • urlopen raises HTTPError on a 4xx. A test that expects a 400 has to catch it and read the body from the exception; tests/test_secret_guard.py's post_status() helper is the pattern.
  • A field has a length limit, and the guard runs before it. A test that pastes an 80-character seed phrase into a 60-character label field gets a secret refusal, not a length refusal. Use a WIF key (52 characters) or a BIP38 key (58) when a test needs a secret that fits a short field, and use the seed phrase where the field has room. Both sides are pinned in tests/test_secret_guard.py.
  • A required control inside a display:none container still blocks submission. The browser refuses to submit and logs "An invalid form field is not focusable". Hiding a section is not enough: the required flags must be cleared, or the section must render something other than required fields. That is why renderSecurityChecks() shows an empty note instead of cards when a key could not be checked, why syncUncheckableState() clears the flags on the wallet software and the compatibility select — also for a cloud key, whose grid is hidden — and why syncRequiredFields() clears the flag on hardware_works and seed_backup_verified. The rotation fields follow the same rule: the type is disabled until the key is marked for rotation, the path is disabled when the key has a saved check and is not marked, and the script refuses to send an empty path. The login page sidesteps it with two separate forms, so only the visible one is ever validated.
  • The mismatch confirmation follows its warning. #address-path-confirmed is required only while #address-path-warning is visible, so the browser itself stops the submission and a required control is never hidden.
  • The per-field secret reminder is gone; only the setup keeps one. It used to be a <span> built in JavaScript (settings.js put it inside a .field label, where a <p> is not allowed) with display: block and font-weight: 400 so it looked right in both places. Since 32.2 it is a single p.secret-warning in setup.html, and setup.js, health-check.js and settings.js no longer build one at all. A test asserts setup.html holds exactly one class="secret-warning" and that the check form and the settings hold none, so re-adding a per-field copy breaks it.
  • static/health-check.js must not use placeholder anywhere. A test asserts the string is absent, because a previous value is a real input value.
  • The key card's line order is pinned by two test modules. tests/test_key_status.py asserts card.append(top, name) → card.append(model) → card.append(path), and tests/test_health_check_form.py asserts that the top line is ${String(entry.key_number).padStart(2, "0")} · ${keyTypeLabel(entry.type)}, that card.append(type) is gone, and that the model line is skipped by sameLabel(entry.model, entry.name). Adding a line, reordering them, or dropping sameLabel() breaks them.
  • The dashboard heading is dynamic and the static description is gone. tests/test_app.py asserts id="setup-shape" and the fallback text on the dashboard and assertNotIn("Bitcoin Multisig Health Monitor") there, while tests/test_pwa.py still asserts that exact description in the manifest. Do not move the one into the other.
  • The stylesheet test asserts on the shape of the theme, not on its wording. test_stylesheet_is_the_retro_theme checks the two bevel colour pairs, the dotted focus outline, the font stack, the absence of box-shadow and gradient, and that every border-radius in the sheet is 0. It reads the values, so a restyled rule that keeps the palette passes, and a rounded corner does not.
  • The brand mark is an image, and the class name is pinned. IconTests in tests/test_pwa.py asserts that all six pages carry <img class="brand-mark" src="/icon.svg" alt="" aria-hidden="true">, that the old <span class="brand-mark" aria-hidden="true">V</span> is gone, that the sheet holds image-rendering: pixelated, and that place-items: center appears nowhere. It also pins min-height: 100vh, min-height: 100dvh, flex-direction: column and the exact string main { flex: 1 1 auto; }, which is what keeps the footer at the bottom. Do not style the brand mark as a letter again, and do not drop the main rule.
  • The manifest lists the SVG first, and the two PNGs after it. test_the_manifest_lists_the_served_icon reads icons[0], so a PNG added in front of the SVG turns the suite red. The PNG entries must carry the real size (192x192 and 512x512) and type: "image/png", and test_the_png_icons_are_the_size_the_manifest_declares reads each file's IHDR, so a PNG rasterized at the wrong size fails rather than quietly breaking the install prompt. The two SVG files must also keep the same drawing — test_the_served_icon_carries_the_same_drawing_as_the_package_icon compares them from <svg> onwards, and only the leading comment may differ.
  • make x86 runs prettier --check startos, and the config is the SDK's own. .prettierrc is the single line "@start9labs/start-sdk/prettier.config.json", which is { trailingComma: "all", tabWidth: 2, semi: false, singleQuote: true } — everything else is prettier's default, so printWidth is 80 and quoteProps is "as-needed". That last one silently unquotes a single-word i18n key (Username: 21, not 'Username': 21), and 80 columns decides when an array element or an argument collapses onto one line. Write it the way prettier would, or run npx prettier --config node_modules/@start9labs/start-sdk/prettier.config.json --write startos.
  • The page heading uses class="eyebrow" too. A regex over health-check.html finds Annual security check before the numbered sections. settings.html starts with the h1 Settings, then 01 / Wallet software and 02 / Key rotation — the eyebrow that repeated the h1 was removed in round 32.3. tests/test_health_check_form.py asserts those two lists exactly, so a new eyebrow on either page breaks it, and so does a removed one: update the list in the same change. The 32-B warnings deliberately used class="secret-warning", not an eyebrow.
  • A static help line that the script overwrites is dead text. #software-help in health-check.html was emptied in round 32.3, because lockSoftwareFields() fills it before the user can read it. static/health-check.js must keep setting it — the wording depends on the selected key. A test pins both halves: the empty element in the markup and softwareHelp.textContent in the script.
  • aria-labelledby may point at the eyebrow. login.html's first form has no h2 any more; its aria-labelledby="login-form-title" points at the eyebrow <p> that carries the same id. Do not add a heading back just to have one, and do not drop the id.
  • Do not assert on the quoting style of the JavaScript. static/health-check.js uses single quotes for constant selectors and template literals where a ${...} substitution is needed, and static/settings.js uses double quotes. Assert on the selector text, e.g. [name="address_path"], and on quote-free substrings such as /api/settings and /settings?rotate=.
  • clean_record() in tests/test_alerts.py has a fixed provider list and carries key_id and wallet_software. A test that needs another provider must replace the entry itself, and the recorded software is Sparrow Wallet, which is what a software alert names. A test that exercises the cloud-key form of software_incompatible has to pass cloud_names to build_alerts(). The alert texts are asserted there, and not in tests/test_app.py, which checks the code, the severity, the source, the actions and the key_id only — so a message change touches tests/test_alerts.py alone. A security alert whose provider is the key's own name carries the key source, so a test that looks an alert up by source must use Key N: <name> for that form and the plain software name for the other.
  • sample_payload() / unavailable_payload() are duplicated in tests/test_validation.py, tests/test_database.py, tests/test_app.py, tests/test_maintenance.py and tests/test_secret_guard.py. They build the labels from the key's name and the selected software, they always send contact_info_verified, and for a hardware key hardware_works. They must send the key's own address_path, because the server compares it with the stored one.
  • Every test fixture uses one address path per key (ADDRESS_PATH, SECOND_ADDRESS_PATH, THIRD_ADDRESS_PATH), because the setup refuses duplicates. tests/test_app.py also has ROTATION_PATHS and the rotate() helper, which picks an unused path per key so a test can rotate several keys in one run.
  • validate_health_check() takes a third argument (the wallet software list), and validate_rotation() takes a fourth for the same reason. tests/test_validation.py and tests/test_database.py wrap the first one in a small validate() helper.
  • tests/test_app.py has seven test classes. ServerTestCase starts the server and keeps the session and CSRF cookies a sign-in returns; AppHandlerTests, RenamedSetupTests, SettingsApiTests, RotationApiTests and AcknowledgementApiTests store the default or renamed setup and call authenticate(); AuthApiTests starts from a database with no admin and covers the gating, the first-run admin, the login and logout, and the CSRF check; and KeySetupApiTests creates its admin through POST /api/admin and otherwise starts from an empty database. A new test must go in the class whose setup it needs. Every request helper sends the cookies and the CSRF header, and assert_status() is the one place an error status is expected.
  • The other server test modules share tests/server_session.py. test_export_filename.py, test_key_model.py, test_key_status.py, test_multisig_setup.py, test_pwa.py and test_secret_guard.py mix in SessionMixin, which signs in through the API and carries the cookies; each class calls authenticate() at the end of its own setUp. A test that switches DATABASE_PATH to another file has to call authenticate() again, because the session secret lives in the database, so the old cookie is signed with the wrong secret and every request answers 401. The mixin resolves the path with app.get_database_path(), so the admin lands in the file the server actually reads.
  • Section 19 records one key per submission, so a Check id in the export can cover up to three keys only for rows saved before section 19. A legacy record with several key_checks is read through its first one when an alert needs the key it belongs to.
  • The database path default is repeated in sixteen places. database.py carries database_path="data/vardkey.sqlite3" as the default of every storage function (fifteen of them), and app.py carries it in get_database_path(). A missed one makes a single code path read or create the other file, silently. Grep for the old name before and after a rename round.
  • The favicon link sits next to the manifest link, and /login has no manifest. IconTests asserts the exact string <link rel="icon" type="image/svg+xml" href="/icon.svg"> on all six pages, so a reworded or reordered link breaks it. /icon.svg and the two PNGs are open (they are not in PROTECTED_PAGES and do not start with /api/), which is what lets the login page and the browser tab read them.

Product clarification

The key's configured name is its product label, and a cloud key's name is also a wallet-software choice. A hardware key name is never a wallet-software choice, and a cloud key records its own name as its software. A key is identified by its type and its address path, while the name is only a label that the user may change. The address path is unique per key, and a rotation replaces the key with a new one that must carry a new address path and may carry the other type. A rotation creates a new wallet, so the funds have to be moved from the old multisig setup to the new one before the old keys are discarded. An alert is closed by the user with the action that belongs to it, and a key with an open alert cannot be checked until the alert is closed.