Files
Sam Rolfe 4f233a8012 Phase 8: overrides (shared keycode + versioned admin lock/unlock/grant API)
Build on Phases 1–7. Adds the two override paths from AGENTS.md §4 FR-10/FR-11:
the shared device keycode and the versioned admin HTTP API the separate Kotlin
admin app will call.

- store: migration 004 adds state.grant_until and a settings table. New
  overrides.go records every override in `overrides` and verifies the shared
  keycode (bcrypt; one code for every device, §10 rule 3). ExpireLocks also
  clears an expired grant, and NextScheduleBoundary includes the grant end so a
  device re-locks on time.
- effective state: a grant unlocks a device inside a schedule window; a manual
  lock clears a grant and a grant clears a manual lock. The retained config
  payload gains grant_until (additive, v stays 1).
- web: /api/v1/admin/... (list, state, lock, unlock, grant) guarded by
  CONTROL_ADMIN_API_TOKEN, separate from the session login, plus the on-device
  keycode endpoint POST /api/v1/device/override with per-address throttling.
  A state change is published retained before the API responds.
- cmd/control: `keycode` subcommand and start-up seeding from
  CONTROL_DEVICE_KEYCODE_HASH.
- deploy/verify-phase8.sh + `make verify-phase8` prove the definition of done:
  an API lock reaches a real broker subscriber in well under a second.
- Version 0.6.0-phase8 (compose, Makefile, docs).

Measured on .13: the lock reached the TLS broker subscriber in 7 ms (handler
1 ms, request to retained publish accepted).

Tests: go vet ./... and go test -count=1 ./... pass in golang:1.25-alpine.

Co-Authored-By: Paperclip <noreply@paperclip.ing>
2026-10-13 16:27:46 +11:00

234 lines
7.8 KiB
Go

// Overrides and the shared device keycode (AGENTS.md §4 FR-10, §7 overrides,
// §10 rule 3, §11 Phase 8).
//
// An override is a deliberate change to a device's desired state: a lock, an
// unlock, or a grant of time. Every override is written to the `overrides`
// table and, through the caller, to the append-only `audit` log (FR-15). The
// state table holds only the current value; `overrides` is the history.
//
// The device keycode is a single code shared by every device (§10 rule 3). It
// is stored only as a bcrypt hash in `settings`, is seeded from the environment
// at start-up, and can be changed with `control keycode` (or, later, from the
// dashboard).
package store
import (
"context"
"database/sql"
"errors"
"fmt"
"strings"
"time"
"golang.org/x/crypto/bcrypt"
)
// keycodeSettingKey is the settings row that holds the shared device keycode.
const keycodeSettingKey = "device_keycode_hash"
// ErrNoKeycode is returned when no device keycode has been set.
var ErrNoKeycode = errors.New("device keycode is not set")
// Override is one row of the override history.
type Override struct {
ID int64
DeviceID int64
Actor string
Action string
Until *time.Time
CreatedAt time.Time
}
// insertOverride appends one override row using the given executor, so it can
// be written atomically with the state change it records.
func insertOverride(ctx context.Context, ex interface {
ExecContext(context.Context, string, ...any) (sql.Result, error)
}, deviceID int64, actor, action string, until *time.Time) error {
var untilArg any
if until != nil {
untilArg = until.UTC().Format(time.RFC3339)
}
if _, err := ex.ExecContext(ctx,
`INSERT INTO overrides (device_id, actor, action, until) VALUES (?, ?, ?, ?)`,
deviceID, actor, action, untilArg); err != nil {
return fmt.Errorf("override insert: %w", err)
}
return nil
}
// ListOverrides returns a device's override history, newest first.
func (s *Store) ListOverrides(ctx context.Context, deviceID int64, limit int) ([]Override, error) {
if limit <= 0 || limit > 500 {
limit = 50
}
rows, err := s.db.QueryContext(ctx, `
SELECT id, device_id, actor, action, until, created_at
FROM overrides WHERE device_id = ?
ORDER BY id DESC LIMIT ?`, deviceID, limit)
if err != nil {
return nil, fmt.Errorf("list overrides: %w", err)
}
defer rows.Close()
out := make([]Override, 0, limit)
for rows.Next() {
var (
o Override
until sql.NullString
created sql.NullString
)
if err := rows.Scan(&o.ID, &o.DeviceID, &o.Actor, &o.Action, &until, &created); err != nil {
return nil, fmt.Errorf("scan override: %w", err)
}
o.Until = parseTimePtr(until)
o.CreatedAt = parseTime(created.String)
out = append(out, o)
}
if err := rows.Err(); err != nil {
return nil, fmt.Errorf("list overrides: %w", err)
}
return out, nil
}
// SetDeviceGrant gives a device a time-limited exemption: it is unlocked until
// `until`, even if a weekly window is active. It clears any manual lock, writes
// one override row and one audit row in the same transaction, and returns the
// fresh device. `until` must be in the future.
func (s *Store) SetDeviceGrant(ctx context.Context, id int64, until time.Time, actor, action string) (*Device, error) {
if until.IsZero() {
return nil, errors.New("grant needs an end time")
}
until = until.UTC()
if !until.After(time.Now()) {
return nil, errors.New("grant end time must be in the future")
}
if _, err := s.GetDevice(ctx, id); err != nil {
return nil, err
}
tx, err := s.db.BeginTx(ctx, nil)
if err != nil {
return nil, fmt.Errorf("set grant: %w", err)
}
defer func() { _ = tx.Rollback() }()
if _, err := tx.ExecContext(ctx, `
INSERT INTO state (device_id, locked, until, grant_until, updated_at)
VALUES (?, 0, NULL, ?, datetime('now'))
ON CONFLICT(device_id) DO UPDATE SET
locked = 0,
until = NULL,
grant_until = excluded.grant_until,
updated_at = datetime('now')`,
id, until.Format(time.RFC3339)); err != nil {
return nil, fmt.Errorf("set grant: %w", err)
}
if err := insertAudit(ctx, tx, actor, action, &id,
"granted time until "+until.Format(time.RFC3339)); err != nil {
return nil, err
}
if err := insertOverride(ctx, tx, id, actor, action, &until); err != nil {
return nil, err
}
if err := tx.Commit(); err != nil {
return nil, fmt.Errorf("commit grant: %w", err)
}
return s.GetDevice(ctx, id)
}
// ---------- device keycode ----------
// SetKeycodeHash stores the bcrypt hash of the shared device keycode and
// appends one audit row. The hash is validated first, so a corrupt value never
// replaces a working code.
func (s *Store) SetKeycodeHash(ctx context.Context, hash, actor string) error {
if _, err := bcrypt.Cost([]byte(hash)); err != nil {
return fmt.Errorf("device keycode hash is not a valid bcrypt hash: %w", err)
}
tx, err := s.db.BeginTx(ctx, nil)
if err != nil {
return fmt.Errorf("set keycode: %w", err)
}
defer func() { _ = tx.Rollback() }()
if _, err := tx.ExecContext(ctx, `
INSERT INTO settings (key, value, updated_at) VALUES (?, ?, datetime('now'))
ON CONFLICT(key) DO UPDATE SET value = excluded.value, updated_at = datetime('now')`,
keycodeSettingKey, hash); err != nil {
return fmt.Errorf("set keycode: %w", err)
}
if err := insertAudit(ctx, tx, actor, "keycode.updated", nil, "shared device keycode changed"); err != nil {
return err
}
if err := tx.Commit(); err != nil {
return fmt.Errorf("commit keycode: %w", err)
}
return nil
}
// KeycodeHash returns the stored bcrypt hash, or ErrNoKeycode.
func (s *Store) KeycodeHash(ctx context.Context) (string, error) {
var value string
err := s.db.QueryRowContext(ctx, `SELECT value FROM settings WHERE key = ?`, keycodeSettingKey).Scan(&value)
if errors.Is(err, sql.ErrNoRows) || value == "" {
return "", ErrNoKeycode
}
if err != nil {
return "", fmt.Errorf("keycode hash: %w", err)
}
return value, nil
}
// VerifyKeycode reports whether code matches the stored shared keycode. A
// missing or invalid stored hash never matches, so a broken settings row cannot
// unlock a device by accident.
func (s *Store) VerifyKeycode(ctx context.Context, code string) bool {
code = strings.TrimSpace(code)
if code == "" {
return false
}
hash, err := s.KeycodeHash(ctx)
if err != nil {
return false
}
return bcrypt.CompareHashAndPassword([]byte(hash), []byte(code)) == nil
}
// SeedKeycodeHash stores hash only when no code is stored yet. The Control
// Service calls it at start-up so CONTROL_DEVICE_KEYCODE_HASH in environment.d
// is the first-run source of the code. Once a code exists, the settings row is
// authoritative: a restart never overwrites a code changed with `control
// keycode`, so a runtime change is durable. It returns whether it stored a code.
func (s *Store) SeedKeycodeHash(ctx context.Context, hash string) (bool, error) {
hash = strings.TrimSpace(hash)
if hash == "" {
return false, nil
}
if _, err := bcrypt.Cost([]byte(hash)); err != nil {
return false, fmt.Errorf("CONTROL_DEVICE_KEYCODE_HASH is not a valid bcrypt hash: %w", err)
}
if _, err := s.KeycodeHash(ctx); err == nil {
return false, nil // already set; never overwrite at start-up
} else if !errors.Is(err, ErrNoKeycode) {
return false, err
}
tx, err := s.db.BeginTx(ctx, nil)
if err != nil {
return false, fmt.Errorf("seed keycode: %w", err)
}
defer func() { _ = tx.Rollback() }()
if _, err := tx.ExecContext(ctx, `
INSERT INTO settings (key, value, updated_at) VALUES (?, ?, datetime('now'))
ON CONFLICT(key) DO UPDATE SET value = excluded.value, updated_at = datetime('now')`,
keycodeSettingKey, hash); err != nil {
return false, fmt.Errorf("seed keycode: %w", err)
}
if err := insertAudit(ctx, tx, "system", "keycode.updated", nil, "seeded from CONTROL_DEVICE_KEYCODE_HASH"); err != nil {
return false, err
}
if err := tx.Commit(); err != nil {
return false, fmt.Errorf("commit seeded keycode: %w", err)
}
return true, nil
}