The 13 checks — TennoTools Profile Check

How a check works

fetch → parse → validate → verdict

One HTTP request pulls the profile; @wfcd/profile-parser parses it; the 13 checks run through a safe wrapper (an exception becomes a failed check); the verdict is derived from critical flags only. Details in the guide.

Each run is persisted in data/app.sqlite — summary, individual checks, request metadata.

The checks

6 critical / 7 advisory

#	Check (exact label)	Critical
1	`profile.displayName is a non-empty string`	yes
2	`profile.accountId matches the requested id`	no
3	`profile.accountId is a 24-char hex ObjectId`	yes
4	`profile.masteryRank is an integer >= 0`	yes
5	`profile.created is a valid Date`	yes
6	`profile.loadout is built (suits/xpInfo arrays)`	yes
7	`profile.missions is an array`	no
8	`profile.challengeProgress is an array (Nightwave)`	no
9	`profile.intrinsics is an object`	no
10	`stats.timePlayedSec is a finite number`	yes
11	`stats.missionsCompleted is a number`	no
12	`stats.weapons and stats.enemies are arrays`	no
13	`xpCacheExpiryDate / ceremonyResetDate are parsed Dates`	no

The verdict

derived from critical flags only

broken warnings ok

broken — at least one critical check failed: the parser pipeline is unusable for its purpose.
warnings — only non-critical checks failed: the profile still parses, but a field drifted.
ok — all 13 checks passed.

Keeping the labels honest

frozen by tests

The exact labels and critical flags above are asserted by the test suite — they cannot drift silently.
An offline parity test replays a real captured DE response (packages/core/test/fixtures/profile.raw.json) through the same pipeline on every npm test — no network.
Refresh the fixture with npm run check:live (exactly one request), or run the whole pipeline offline with npm run check:offline.