27 KiB
AI agent note — VardKey
Short working handover for the next model. Keep it short.
docs/spec.mdis the specification (English). §1–§18 are binding and describe the app as it is now, §19–§33 are history as one line per round, and §32 holds nothing. The section numbers are stable — the code, the tests and the documents cite them — so a section is never renumbered and never reused.- The full reasoning for every round is in the git history, as are the earlier Norwegian revision of the specification and the deleted bug report. Do not recreate either file.
- This note carries only what the spec does not: the status, the way to work in this repository, and the traps that have already cost time. Do not repeat the spec here.
Status
- Done and verified: everything in §1–§33. Schema version 13. 534 green tests. The StartOS package is built, sideloaded and running on a StartOS 0.4.0 box.
- The package version is
2026.10.28:0, tagv2026.10.28_0. The version scheme is date based,YYYY.MM.DD:revision;UPDATING.mdcarries the rule. - What is left: nothing in the spec. Two things only the user can close are listed at the top of
docs/spec.md: the box and backup tests have to be re-run before the first release, and the prohibition lists are re-confirmed before release. - 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 atC:\Users\<brukarnamn>\Downloads\vardkey, so they must pass there as well as on Linux.
Working in this repository
- The Gitea tools write a whole file at a time (
gitea_update_file, auto-SHA, one file per call), and there is no preview step and no shell. There is no branch delete either. - The agent therefore cannot run the suite,
make x86,tsc --noEmitorprettier --check. The user runs them, and the user is the one who rasterizes an image or deletes a branch by hand. - Keep the documents short.
docs/spec.mdand this note are read by every future model, and their size is context the next round cannot spend.
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 unmistakable shapes, so an ordinary note is not refused). 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; a BIP38 key; and a run of twelve or more
consecutive BIP39 words. It returns a short English label such as "a seed phrase", never the value.
- The embedded wordlist is pinned by the suite — 2048 words, alphabetical, 3–8 lower-case letters, and no two words sharing their first four letters. That last check is the strong one. Do not edit the list without that test.
- The refusal must stay a
ValidationError(SecretDetected), so the existing handlers answer400.app.pyneeds no handler for it. A plainValueErrorwould be caught bySTORAGE_ERRORSand answered500. - The guard runs before the field's length limit. A seed phrase pasted into a 60-character label is answered with "looks like a seed phrase", not with "too long". This order is deliberate and pinned by tests; do not swap it back.
- The UI reminder is one line, in the key setup, and nowhere else; the page notice is a different
thing. A test pins exactly one
class="secret-warning"insetup.htmland none in the other two pages. - The limitations are 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); only the English wordlist is used; a secret split across several fields, or in a shape outside the patterns, is not caught; and a note that happens to be twelve or more BIP39 words in a row is a false positive.
Alerts and key identity — the parts that are easy to get wrong
The rules are in docs/spec.md §7; these are the mechanisms that have already cost time.
- The acknowledgement stores the key.
alert_acknowledgements.key_idis nullable and holds the key the alert belonged to, orNULLfor a software alert raised by the settings list. An alert with a key is hidden by(code, key_id); one without a key stays on(code, source). This closed a real bug: a rename used to reopen a closed alert, becausesourcecarries the name the key has now. - The row in
ai_security_checksdecides the source, never the configured names._security_alerts()looks the row's ownkey_idup in a lookup built bybuild_alerts(), so a rename or a rotation cannot turn a key alert into a software one. build_alerts()andcollect_alerts()takecloud_names=(), and every caller must pass the same tuple. It is threaded throughrotation_alert_targets(),software_alert_targets(),newest_alert_check_id()and the four call sites inapp.py, andapp.cloud_key_names(keys)builds it. Ifnewest_alert_check_id()sees different names, the alert'ssourcediffers, the acknowledgement storescheck_id = None,_is_acknowledged()refuses to hide an alert without a check id, and the alert can never be closed. Keep the default()onbuild_alerts()._send_health_check_list()passes the same tuple forneeds_follow_upandfollow_up_open.software_incompatiblehas two forms. A software from the settings list keepssource= the software name,key_id=None, actions[remove_software], and carriessoftware. 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 nosoftwarevalue. Only that form names the software in its message, because its source is the key.- Every closure carries its source, because the two forms of a security code are otherwise
indistinguishable: the CSV writes
<source> — <code>: <action> (YYYY-MM-DD), and the history table showsSource,Alert,ActionandClosed. needs_follow_upis frozen andfollow_up_openis current. Do not "fix"needs_follow_upto followcollect_alerts(), which hides acknowledged alerts and would rewrite history.- Deliberately left out: a software-sourced
security_vulnerableoffersremove_softwarenext toignore(one button leads nowhere), and a software-sourcedsecurity_compromisedoffers it as its only action, so an alert whose "software" is a cloud key name can never be closed. Neither is fixed.
Data model and migration (schema version 13)
The tables and columns are in docs/spec.md §4. What matters here:
- Startup migration rebuilds
keyswhen it still hasdevice, the oldkey_numberCHECK, or noactiveflag;key_checkswhen it uses the old layout or still holds the two cloud columns; andhealth_checkswhen itsCHECKconstraints still allow the old values. It then dropstool_checksif it is there, adds the newer columns throughOPTIONAL_COLUMNS, backfills the key snapshot on every check from the key it belongs to, and creates the partial unique index over the active row. - The table definitions are
{table}templates, so a rebuild can create the same table under a temporary name.health_checksis a parent table, so the migration runs withPRAGMA foreign_keys = OFF, checksPRAGMA foreign_key_checkbefore committing, and rolls back on any failure. tests/test_database.pycarriesLEGACY_SCHEMAandSCHEMA_VERSION_5,_8,_9,_10,_11and_12as literal older schemas. Copy that pattern for a new migration test: write the old schema, runinitialize_database(), then assert on the migrated rows.- There is no data migration for an older test database: the user deletes the
datafolder and configures the setup again.
API
The endpoint list and the field details are in docs/spec.md §11 and docs/application.md. Six
things that are easy to break:
- New static files must be added to
STATIC_FILESinapp.py, or they return404./robots.txtistext/plain; charset=utf-8and is deliberately absent from the service worker's shell list. STORAGE_ERRORSinapp.pyincludesRuntimeError, becauselist_keys()raises it for a partial setup, andValueError. Keep both, or a broken database answers with a traceback instead of a500. It is also whyvalidate_admin_credentials()is called in its owntry, and why a refused secret has to be aValidationErrorrather than a bareValueError.DuplicateAddressPathandWalletSoftwareNameInUsemust be caught before the generalValidationError— in_save_key_settingsand_rotate_key— or the409turns into a400.newest_alert_check_id()derives the check id for the acknowledgement, the rotation and the software removal, so all three use the same comparison rule.blocking_alert()is the module-level helper the409is built on, and_blocking_alert()reads the key id straight from the payload, so the rule applies before validation.- The rotate payload is
replaced+keep_name+name+type+address_path, andtypeis optional. A rotation inserts a new row and retires the old one, and the acknowledgements are written against the old key id. - The
#history-detailanchor and the?id=query on/historyare client-side only: they never reach the server, so no route changed for them.
The StartOS package — the parts that bite
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 application and the package share one repository, at the root. 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.mdstates the deviation so nobody adds anupstream-project/submodule to "fix" it. - Anything the
Dockerfiledoes notCOPYis not in the image. The modules go in withCOPY *.py ./and the assets withCOPY static ./static.static/must be copied:app.pyanswers/healthfrom code, but every page is read fromROOT/static, so an image without it reports a healthy service whose whole UI answers404. - Three things must agree.
startos/utils.tsholdsuiPort(8080);startos/main.tspasses it asPORTand uses it in the readiness URL;startos/interfaces.tsbinds it. A redECONNREFUSED 127.0.0.1:8080blip beforeVardKey listening on port 8080is just the first poll landing before Python has bound the socket; nothing to fix. - Two actions, registered through
Actions.of()instartos/actions/index.ts, with their strings instartos/i18n/dictionaries/:delete-all-dataandset-admin-password. Both runmaintenance.pyin awithTempsubcontainer, behind a confirmation the handler enforces, and the generated password never travels into the container. - The admin task in
startos/init/requireAdmin.tsraises and clears itself, because SDK 3.0.4 has no condition for "the admin password is unset", and its probe fails open. On the box it shows on Tasks as recommended and does not hold the service back — the app's own login gate is what refuses everything.README.mddocuments that behaviour. - Backups.
startos/backups.tsissdk.Backups.ofVolumes('main'). StartOS stops the service for the duration of a backup, so the database file is quiescent while it is copied — no snapshot hook is needed. - 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.s9pkis built locally withmake. - The skeleton 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-sdkis pinned to3.0.4.node_modules/is not in the repository. - The root
icon.svgis the package artwork, read from the repository root by the build, the same retro drawing asstatic/icon.svg, and it must stay under 40 KiB. - Documentation layout.
README.mdis the StartOS package README, with the guide's fixed heading set.docs/spec.mdis the specification,docs/application.mdthe application and API reference,instructions.mdthe user quick start, andAGENTS.mdthe repo-specific agent context (CLAUDE.mdis the one-line@AGENTS.mdimport). Development happens onhttps://gitea.sondreolav.com/ai-bot/vardkey(theoriginremote), andhttps://github.com/sondreolav/vardkeyis its public mirror;AGENTS.mdsays which links point where and why. Do not point the manifest or the README links back at Gitea.
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 andTemporaryDirectory.cleanup()fails withWinError 32. Always close explicitly withtry/finally; seewrite_schema()intests/test_database.py.- A module the image is missing is a runtime error, not a build error.
auth.pywas once left out of theDockerfile, and bothapp.pyandmaintenance.pythen died withModuleNotFoundErrorinside the container while every test stayed green. TheCOPY *.py ./line is what stops that recurring. - 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, andurllibfollows it, so an assertion on page markup reads the login page's HTML instead of failing with401. Usetests/server_session.py'sSessionMixin. The manifest, the icons and/robots.txtare open while the pages are not. urlopenraisesHTTPErroron a 4xx. A test that expects a400has to catch it and read the body from the exception;tests/test_secret_guard.py'spost_status()helper is the pattern. A429is read the same way.- A field has a length limit, and the guard runs before it. Use a WIF key (52 characters) or a BIP38 key (58) when a test needs a secret that fits a short field, and the seed phrase where the field has room.
- A
requiredcontrol inside adisplay:nonecontainer still blocks submission. Hiding a section is not enough: the required flags must be cleared. That is whyrenderSecurityChecks()shows an empty note instead of cards when a key could not be checked, whysyncUncheckableState()clears the flags on the wallet software and the compatibility select, and whysyncRequiredFields()clears the flag onhardware_worksandseed_backup_verified. The rotation fields follow the same rule: the type isdisableduntil the key is marked for rotation, the path isdisabledwhen the key has a saved check and is not marked, and the script refuses to send an empty path. - A hidden element has no box, so an anchor cannot jump to it.
#history-detailis hidden until a check is opened, sorevealDetail()inhistory.jsunhides the section first and only then callsscrollIntoView(). Do not reorder those two lines. - A bare substring assertion over page markup also catches comments. Assert on the element, not on the phrase.
- A
display: gridpanel ignores thehiddenattribute..menu-panel[hidden] { display: none; }is in the sheet for that reason, the same way.form-stack[hidden],.field-grid[hidden]and.form-section[hidden]already were. - The menu's outside-click listener has to allow the button's own click. The button's handler runs
first and toggles the panel; the document listener then sees a target inside
.menuand does nothing. - The current-page check compares a normalised path, and the current item is a span.
currentPath()strips trailing slashes and keeps the root as/. The match is rendered as a<span class="menu-item menu-item-current">witharia-current="page", never an<a>. - A sticky top bar needs a scroll container that is the viewport.
.topbaris a flex child ofbody, andbodysets nooverflow, soposition: sticky; top: 0works — but it stops working the moment an ancestor getsoverflow: hiddenorauto. - The mismatch confirmation follows its warning.
#address-path-confirmedisrequiredonly while#address-path-warningis visible, so the browser itself stops the submission. static/health-check.jsmust not useplaceholderanywhere. 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.pyassertscard.append(top, name)→card.append(model)→card.append(path), andtests/test_health_check_form.pyasserts the top line, thatcard.append(type)is gone, and that the model line is skipped bysameLabel(entry.model, entry.name). - The dashboard heading is dynamic and the static description is gone.
tests/test_app.pyassertsid="setup-shape"and the fallback text on the dashboard, whiletests/test_pwa.pystill asserts the manifest description. Do not move the one into the other. - The stylesheet test asserts on the shape of the theme, not on its wording. It checks the two
bevel colour pairs, the dotted focus outline, the font stack, the absence of
box-shadowandgradient, and that everyborder-radiusin the sheet is0. - The brand mark is an image, and the class name is pinned.
IconTestsintests/test_pwa.pyasserts that all six pages carry<img class="brand-mark" src="/icon.svg" alt="" aria-hidden="true">, that the sheet holdsimage-rendering: pixelated, and thatplace-items: centerappears nowhere. It also pinsmin-height: 100vh,min-height: 100dvh,flex-direction: columnand the exact stringmain { flex: 1 1 auto; }. Do not style the brand mark as a letter again. - The two SVG files must keep the same drawing. The package icon at the repository root is the
source of truth, and
test_the_served_icon_carries_the_same_drawing_as_the_package_iconcompares them from<svg>onwards. A change to one drawing has to be made in both. - The two PNGs are real files, rasterized by the user, because the agent has no rasterizer.
IconTestsreads each file's IHDR and asserts 192×192 and 512×512. - The manifest lists the SVG first, and the two PNGs after it.
test_the_manifest_lists_the_served_iconreadsicons[0]. The PNG entries must carry the real size andtype: "image/png". - The favicon link sits next to the manifest link, and
/loginhas no manifest.IconTestsasserts the exact string<link rel="icon" type="image/svg+xml" href="/icon.svg">on all six pages. - The prettier config is the SDK's own, and
make x86does not run the check..prettierrcis 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, soprintWidthis 80 andquotePropsis"as-needed". The SDK makefile defines the check asFORMAT_CHECK, butmake x86packs without it, so write the TypeScript the way prettier would and runnpx prettier --config node_modules/@start9labs/start-sdk/prettier.config.json --check startosbefore you hand the round back;--writefixes the files. Instartos/manifest/i18n.tsa value that does not fit the line breaks after the colon, and a string with an apostrophe uses double quotes. A property is measured together with its trailing comma, so the limit isprintWidth − indent— 78 columns for a top-level property, not 80.es_ESat 79 columns must break;en_USat 74 andpl_PLat 76 stay inline. - The page heading uses
class="eyebrow"too.tests/test_health_check_form.pyasserts the eyebrow lists ofhealth-check.htmlandsettings.htmlexactly, so a new eyebrow breaks it, and so does a removed one. The secret warnings deliberately useclass="secret-warning", not an eyebrow. - A static help line that the script overwrites is dead text.
#software-helpinhealth-check.htmlis empty, becauselockSoftwareFields()fills it before the user can read it. A test pins both halves: the empty element in the markup andsoftwareHelp.textContentin the script. aria-labelledbymay point at the eyebrow.login.html's first form has noh2; itsaria-labelledby="login-form-title"points at the eyebrow<p>that carries the sameid.- Do not assert on the quoting style of the JavaScript.
static/health-check.jsuses single quotes for constant selectors and template literals where a${...}substitution is needed, andstatic/settings.jsuses double quotes. Assert on the selector text and on quote-free substrings. - A test string that spans a line break needs the escape, not a literal backslash. Two assertions
in
tests/test_health_check_form.pyonce carried\\nwhere the script has a real newline, so they looked for a backslash and a letter and failed. Prefer an assertion that names the two halves on either side of the break, or normalise with" ".join(script.split()). clean_record()intests/test_alerts.pyhas a fixed provider list and carries the key on the security rows. Eachai_security_checksentry is a dict withprovider,key_idandstatus: the entry whose provider is the key's own name carries the key number, and the others carryNone. The alert texts are asserted there, and not intests/test_app.py, which checks the code, the severity, the source, the actions and thekey_idonly.sample_payload()/unavailable_payload()are duplicated in five test modules. They must send the key's ownaddress_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).tests/test_app.pyalso hasROTATION_PATHSand therotate()helper. With schema 13 a rotation retires the key row, so a test that rotates and then records a check has to use the new key id —rotate()hands it back underkey_id. validate_health_check()takes a third argument (the wallet software list), andvalidate_rotation()takes a fourth for the same reason.tests/test_app.pyhas seven test classes. A new test must go in the class whose setup it needs:ServerTestCase,AppHandlerTests,RenamedSetupTests,SettingsApiTests,RotationApiTests,AcknowledgementApiTests,AuthApiTestsandKeySetupApiTests. Every request helper sends the cookies and the CSRF header, andassert_status()is the one place an error status is expected.- The other server test modules share
tests/server_session.py. They mix inSessionMixin, which signs in through the API, and each class callsauthenticate()at the end of its ownsetUp. A test that switchesDATABASE_PATHto another file has to callauthenticate()again, because the session secret lives in the database. - A login lockout test has to clear the counter.
AuthApiTestsshares one database per test, and the lockout lives inapp_settings, so a test that trips it leaves the account locked for the tests that follow unless it resets the two rows or uses its own database file. - The export link and its note live on the history page, and four tests moved with them.
tests/test_export_filename.pyasserts the dashboard has no/api/exportlink;tests/test_app.py'stest_the_dashboard_carries_no_navigation_link_of_its_own;test_the_history_page_offers_the_export_in_its_own_sectionintests/test_health_check_form.pypins theExportsection, the title bar, the note under the button and the absent.page-heading > .secondary-buttonrule; andtest_the_history_links_to_the_detail_sectionpins both anchors. #key-countstill exists insetup.html, but not on the dashboard. The two are on different pages and were never the same element, so assert per file./api/previous-valuesstill returns asecuritymap that nothing consumes any more, and itsaddress_pathper key is no longer used by the form, which pre-fills the path fromkeys.address_path. Both are deliberate.delete_all_data()deletes the checks before the keys.key_checks.key_idisON DELETE RESTRICT, so the other order fails. The acknowledgements are deleted explicitly, because they have no foreign key at all, and the delete has to clear every row ofkeys, active and retired alike.- The PWA cache stores only a valid shell answer.
isCacheable()instatic/sw.jsrequiresresponse.ok,response.type === "basic"and!response.redirected. The allow-list is the other half: the worker never names/api/*,/health,/loginor/robots.txt. - Section 19 records one key per submission, so a
Check idin the export can cover up to three keys only for rows saved before section 19. - The database path default is repeated in sixteen places.
database.pycarriesdatabase_path="data/vardkey.sqlite3"as the default of every storage function (fifteen of them), andapp.pycarries it inget_database_path(). A missed one makes a single code path read or create the other file, silently.
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 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. A rotation also retires the old key row: the multisig slot number stays, the old checks keep the name and type they were recorded with, and the new row starts with an empty history.