Files
nanoclaw/setup/migrate-v1.ts
T
gabi-simons 3ee7d2147e feat: add v1 → v2 migration to setup flow (experimental)
`bash nanoclaw.sh` detects a v1 install before channel pairing and does a
best-effort automated port of operationally important state. Hands off to
a new `/migrate-from-v1` skill for owner seeding and fork customizations.

Between the timezone and channel steps, `setup/auto.ts` calls
`runMigrateV1()` which orchestrates these registered sub-steps (each a
separate entry in the progression log with its own raw log + status
block — failures never abort the chain):

- **migrate-detect** — scans siblings of the v2 checkout + common $HOME
  locations; `$NANOCLAW_V1_PATH` overrides authoritatively. Relaxed
  `package.json` check lets forks + partial installs still match; DB
  presence is the strongest signal.
- **migrate-validate** — asserts v1 DB shape (tables + required
  columns); writes `schema-mismatch.json` on failure. Subsequent steps
  short-circuit their DB-dependent parts but still run.
- **migrate-db** — seeds `agent_groups` + `messaging_groups` +
  `messaging_group_agents` from v1's `registered_groups`. JID
  decomposition (`dc:123` → `channel_type='discord'`,
  `platform_id='discord:123'`); `trigger_pattern` + `requires_trigger`
  → `engage_mode` + `engage_pattern` (mirrors migration 010 backfill).
  Users + user_roles are NOT seeded — the skill does that with an owner
  interview. Idempotent: existing rows reused, not duplicated.
- **migrate-groups** — rsync group folders. v1 `CLAUDE.md` → v2
  `CLAUDE.local.md` (v2 composes `CLAUDE.md` at container spawn); v1
  `container_config` JSON → `.v1-container-config.json` sidecar for the
  skill to translate. Tight v1-pattern scan (`/workspace/ipc/tasks`,
  `store/messages.db`, `[PR_CONTEXT:`, etc.) flags files referencing
  v1-specific infrastructure — content is NOT modified, just flagged in
  the handoff.
- **migrate-env** — merges v1 `.env` into v2 `.env`, never overwriting
  existing v2 keys.
- **migrate-channel-auth** — per-channel registry tracks v1 env keys,
  v2 required keys (with source-of-key instructions — e.g. Discord
  needs `DISCORD_PUBLIC_KEY` which v1 never stored), and candidate
  on-disk auth state paths (Baileys keystore, matrix sync state,
  etc.). Missing required v2 keys surface as actionable followups and
  flip the step to `partial`.
- **migrate-channels** — runs `setup/install-<channel>.sh` for each
  detected channel in non-interactive mode. Install-script output is
  captured to `logs/setup-migration/install-<channel>.log` sidecars
  (silent under the parent spinner). Channels with no v2 adapter get
  a `not_supported` followup but don't degrade status.
- **migrate-tasks** — v1 `scheduled_tasks` → `messages_in` rows with
  `kind='task'` in each session's `inbound.db`. `schedule_type`
  mapping (cron / interval / once → v2 cron). Idempotent: skips v1
  task ids already present. Inactive rows dumped to
  `inactive-tasks.json` for reference.

Everything writes to `logs/setup-migration/handoff.json` — the source
of truth the skill consumes.

`.claude/skills/migrate-from-v1/SKILL.md`:

- **Phase A** (always): owner seeding + v1 access policy flip
  (`unknown_sender_policy` public/strict) via `AskUserQuestion`. Pulls
  sender candidates from v1's `messages` table as hints.
- **Phase B** (if followups exist): walks
  `handoff.followups` — translates `.v1-container-config.json`
  sidecars, handles `not_supported` channels, fills in missing
  required keys with instructions on where to get them.
- **Phase C** (fork-aware): `git log <upstream>..HEAD` in v1. Empty →
  "no customizations to port." Non-empty → scope choice (mechanical /
  full interview / reference-only). Portable categories
  (`container/skills/*`, `.claude/skills/*`, docs) scan+copy with
  `scanForV1Patterns`. Non-portable (`src/*`,
  `container/agent-runner/src/*`) stash to `docs/v1-fork-reference/`
  — explicit "don't translate v1 infra to v2" warning because v1's
  IPC file queue / single DB don't exist in v2.

Clearly marked in README, CLAUDE.md, SKILL.md header, and via a `p.warn`
that fires once per run when v1 is detected. Users with no v1 install
see a silent skip — no prompts, no noise.

Verified end-to-end against a live v1 install (300 discord + 1
discord-supervisor groups, fork with ~15 commits of PR-factory work):
- Detect → validate → db (301 rows seeded) → groups (301 CLAUDE.local.md
  + 178 other files + 1 container_config sidecar) → env (4 keys copied)
  → channel-auth (flagged missing `DISCORD_APPLICATION_ID` +
  `DISCORD_PUBLIC_KEY`) → channels (discord installed, discord-supervisor
  → not_supported) → tasks (0 rows, skipped)
- Idempotent re-run: 0 rows created, 903 rows reused; tasks skip if
  id already present
- Fresh-user case: silent skip, no prompts, straight to "You're ready!"
- Schema-mismatch case: recorded to `schema-mismatch.json`, chain
  continues

- Unit tests for the pure transforms (`parseJid`,
  `inferChannelType`, `triggerToEngage`, `scanForV1Patterns`,
  `looksLikeV1Install`)
- Validate `requiredV2Keys` for telegram/slack/matrix/teams/webex/
  resend/linear against the actual Chat SDK packages (Discord was
  verified from real error output)
- Widen candidate auth file paths for WhatsApp/Matrix/iMessage based
  on real non-Discord v1 installs once we have some

See docs/v1-to-v2-changes.md for the v1 → v2 architecture diff.
2026-04-23 13:06:14 +00:00

258 lines
9.2 KiB
TypeScript

/**
* v1 → v2 migration orchestrator. Called from setup/auto.ts after the
* timezone step and before the channel step.
*
* Silent happy path: if no v1 install is found, we emit one "skipped" step
* and return. Users on a fresh v2 install never see anything.
*
* When v1 IS found: detect → [confirm] → group-selection prompt → validate
* → db → groups → env → channel-auth → channels → tasks → handoff.
* Every sub-step is a separate entry in the progression log; failures never
* abort the chain (the handoff file records them for the skill to finish).
*
* After everything runs, a one-line note points the user at the
* `/migrate-from-v1` skill.
*/
import fs from 'fs';
import path from 'path';
import * as p from '@clack/prompts';
import Database from 'better-sqlite3';
import k from 'kleur';
import { ensureAnswer, runQuietStep } from './lib/runner.js';
import { wrapForGutter } from './lib/theme.js';
import * as setupLog from './logs.js';
import {
HANDOFF_PATH,
MIGRATION_DIR,
inferChannelType,
readHandoff,
v1PathsFor,
writeHandoff,
} from './migrate-v1/shared.js';
/**
* Count groups in v1's registered_groups, split by whether the channel_type
* can be inferred. Uses the same `inferChannelType` logic as migrate-db so
* the displayed count matches what will actually get seeded. Open-and-close
* because this runs in the orchestrator before migrate-db's child process.
*/
function countV1Groups(v1Root: string): { total: number; wired: number } {
const dbPath = v1PathsFor(v1Root).db;
try {
const db = new Database(dbPath, { readonly: true, fileMustExist: true });
const rows = db
.prepare('SELECT jid, channel_name FROM registered_groups')
.all() as Array<{ jid: string; channel_name: string | null }>;
db.close();
let wired = 0;
for (const r of rows) {
if (inferChannelType(r.jid, r.channel_name)) wired++;
}
return { total: rows.length, wired };
} catch {
return { total: 0, wired: 0 };
}
}
async function askGroupSelection(counts: { total: number; wired: number }): Promise<'all' | 'wired-only' | 'cancel'> {
// Non-interactive escape hatch for CI / re-runs / scripted migrations.
// NANOCLAW_MIGRATE_SELECTION = 'all' | 'wired-only' | 'cancel'.
const envChoice = process.env.NANOCLAW_MIGRATE_SELECTION?.trim();
if (envChoice === 'all' || envChoice === 'wired-only' || envChoice === 'cancel') {
setupLog.userInput('migrate_selection', `${envChoice} (from NANOCLAW_MIGRATE_SELECTION)`);
return envChoice;
}
// Most v1 installs accumulated many orphan folders. Default the user to
// wired-only (the ones we can actually route) — explicit opt-in for "all".
const choice = ensureAnswer(
await p.select({
message: `Found ${counts.total} v1 group folders (${counts.wired} wired to a channel). Which to bring over?`,
options: [
{
value: 'wired-only',
label: `Only the ${counts.wired} wired ones`,
hint: 'recommended — skips orphans',
},
{
value: 'all',
label: `All ${counts.total} folders`,
hint: 'brings dead/orphan folders over too',
},
{
value: 'cancel',
label: 'Skip migration',
hint: "I'll migrate later",
},
],
}),
) as 'all' | 'wired-only' | 'cancel';
setupLog.userInput('migrate_selection', choice);
return choice;
}
/**
* Finalize the handoff record after every sub-step has run. Computes an
* overall status from per-step statuses: anything `failed` → partial;
* anything `partial` → partial; else success.
*/
function finalizeHandoff(): 'success' | 'partial' | 'failed' {
const h = readHandoff();
const statuses = Object.values(h.steps).map((s) => s?.status);
const anyFailed = statuses.includes('failed');
const anyPartial = statuses.includes('partial');
const overall: 'success' | 'partial' | 'failed' = anyFailed
? 'partial' // DB or files may have landed; the skill can pick up the rest
: anyPartial
? 'partial'
: 'success';
h.overall_status = overall;
writeHandoff(h);
return overall;
}
function printHandoffNote(overall: 'success' | 'partial' | 'failed'): void {
const relHandoff = path.relative(process.cwd(), HANDOFF_PATH);
const lines: string[] = [];
if (overall === 'success') {
lines.push(
wrapForGutter(
'Your v1 install has been migrated. Run `/migrate-from-v1` in Claude next — it will seed your owner account and help port any custom code you had.',
4,
),
);
} else {
lines.push(
wrapForGutter(
'Migration finished with some items for a human. Run `/migrate-from-v1` in Claude — it will read the handoff, finish the unfinished steps, and walk through custom code.',
4,
),
);
}
lines.push('');
lines.push(k.dim(` Handoff: ${relHandoff}`));
lines.push(k.dim(` Full log: ${setupLog.progressLogPath}`));
lines.push(k.dim(` Raw logs: ${setupLog.stepsDir}/`));
p.note(lines.join('\n'), 'Migration handoff');
}
export async function runMigrateV1(): Promise<'proceeded' | 'skipped' | 'cancelled'> {
// 0. Ensure migration log dir exists before any sub-step writes to it.
fs.mkdirSync(MIGRATION_DIR, { recursive: true });
// 1. Detect. If nothing obvious, give the user one subtle chance to point
// us at a non-standard path — then accept silently.
const detect = await runQuietStep('migrate-detect', {
running: 'Checking for a previous NanoClaw install…',
done: 'Found a previous install.',
skipped: 'No previous install to migrate.',
});
const v1Found = detect.ok && detect.terminal?.fields.STATUS === 'success';
if (!v1Found) {
// Silent skip — the 99% case is a fresh install with no v1 anywhere.
// Prompting for a custom path on every fresh run is UX noise. Users
// with a v1 at a non-standard location use `NANOCLAW_V1_PATH=<path>
// bash nanoclaw.sh` (documented in README + setup/auto.ts header).
return 'skipped';
}
// 2. Ask the user which groups to bring over.
const h = readHandoff();
if (!h.v1_path) {
// Shouldn't happen — detect set it if v1Found. Guard anyway.
return 'skipped';
}
// Experimental warning — fires only when a v1 install is found, so stock
// v2 users (no v1 to migrate) never see it. Not a blocker; the user can
// still proceed. Skip when NANOCLAW_MIGRATE_SELECTION is set (scripted /
// CI runs have already accepted the risk by defining their selection).
if (!process.env.NANOCLAW_MIGRATE_SELECTION) {
p.log.warn(
wrapForGutter(
'v1 → v2 migration is experimental. Back up your v2 state (data/v2.db, groups/) before continuing. Not recommended for high-stakes production installs — it does a best-effort port and a human still has to finish via /migrate-from-v1.',
4,
),
);
}
const counts = countV1Groups(h.v1_path);
const selection = await askGroupSelection(counts);
if (selection === 'cancel') {
// Mark the handoff so the skill can still see what would have happened.
const ho = readHandoff();
ho.overall_status = 'skipped';
writeHandoff(ho);
return 'cancelled';
}
// 3. Validate — if it fails, subsequent steps will short-circuit the
// DB-dependent parts. Groups + env still run.
await runQuietStep('migrate-validate', {
running: "Checking the v1 database's shape…",
done: 'v1 database looks good.',
failed: "v1 database didn't match what I expected.",
skipped: 'Skipped database validation.',
});
// 4. DB seeding — parameterized by the user's selection.
await runQuietStep(
'migrate-db',
{
running: 'Seeding v2 agents and channels from v1…',
done: 'Seeded v2 database.',
skipped: 'Skipped database seeding.',
failed: "Couldn't seed the v2 database.",
},
['--selection', selection],
);
// 5. Group folders.
await runQuietStep('migrate-groups', {
running: 'Copying group folders…',
done: 'Group folders copied.',
skipped: 'Skipped group-folder copy.',
failed: "Couldn't copy some group folders.",
});
// 6. Env keys.
await runQuietStep('migrate-env', {
running: 'Merging v1 .env into v2 .env…',
done: 'Env keys migrated.',
skipped: 'No env keys to migrate.',
failed: "Couldn't merge .env.",
});
// 7. Non-env channel auth (Baileys keystore, matrix state, etc.).
await runQuietStep('migrate-channel-auth', {
running: 'Copying channel auth files…',
done: 'Channel auth copied.',
skipped: 'No channel auth to copy.',
failed: 'Some channel auth files need attention.',
});
// 8. Install v2 channel adapters for the detected channels.
await runQuietStep('migrate-channels', {
running: 'Installing v2 channel adapters…',
done: 'Channel adapters installed.',
skipped: 'No channels to install.',
failed: 'Some channel adapters need attention.',
});
// 9. Scheduled tasks.
await runQuietStep('migrate-tasks', {
running: 'Porting scheduled tasks…',
done: 'Scheduled tasks ported.',
skipped: 'No scheduled tasks to port.',
failed: 'Some scheduled tasks need attention.',
});
// 10. Finalize + hand off.
const overall = finalizeHandoff();
printHandoffNote(overall);
return 'proceeded';
}