TRUTH engine

Is it actually true? Let's check.

Sign in to try it

Runs cost credits. Ask us for credits.

Example results

What TRUTH returns

Real results, not mock-ups.

Sent

“God helps those who help themselves.”

Standard read

Claim assessment

“God helps those who help themselves.”

Assessed Oct 14, 2026, 4:15 PMtruth-engine/research-v32

Shepherd’s answer

Too early to call: it needs evidence Shepherd can’t check yet

Shepherd’s first look leaned toward this being false, but the evidence doesn’t let it go that far.

What holds it back

Some evidence this needs is outside what Shepherd can check yet

Claim by claim

Too early to call
Claim 1
God helps people only if those people have first helped themselves.
Shepherd held back from a stronger answer on this part.
Why
The absolutist 'only if' conditional is inconsistent with the biblical witness: Ephesians 2:8–9 treats salvation as God’s gift not earned by works ( / –29), and narrative/providential passages — notably Exodus 14:14 ( / , load-bearing), Psalm 121:2 ( / ), and Matthew 6:26 ( / ) — depict God initiating deliverance and provision. Passages showing the Spirit’s help in human weakness (Rom 8:26; / ) and God’s enabling of will and work (Phil 2:13; / ) further undercut a strict rule that locates divine help only after prior human self‑help, though some texts are covenantal or rhetorical and a focused lexical/contextual study would sharpen the scope of this judgment.

Read Shepherd’s full reasoning

This reasoning was written for Shepherd’s first look. Where it sounds surer than the answer above, the answer above is the one Shepherd stands behind.

On the relevant biblical‑exegetical and theological grounds, the absolutist formulation that “God helps people only if those people have first helped themselves” does not square with the canonical witness provided. Key texts portray divine initiative and provision (e.g. Eph 2:8–9; Exod 14:14; Ps 121:2; Matt 6:26; Rom 8:26; Phil 2:13; Rom 8:28), showing God acting for, providing for, and assisting the weak rather than acting only after prior human self‑help. Some cited passages are soteriological, narrative, or rhetorical and the precise semantic range of “help” and the force of an exclusivizing “only if” require dedicated lexical and contextual study, but the aggregate scriptural pattern undermines the claim as a universal theological rule.

From different angles

None of the angles Shepherd checked gives it enough to call this yet.

Submit a claim (async, real engine)

Funded account · browser session only20 submissions / hour per account202 in <1s; the assessment finishes in ~2–10 minutes (deep: longer)

Queues an assessment and returns a job id; poll /status. Resubmitting the same claim joins the same job.

Request body

Field	Type	Meaning
`content`required	`string (1–2000)`	The claim, lyric, line from a sermon, or saying to test. Words that could ask 2–3 different claims are run as every claim they ask; words asking more than 3 run once, as typed.
`claims`	`string[] (1–3)`	Instead of content: the exact claims to test, each run as written.
`context`	`string (≤2000)`	Where it was said / who said it.
`reasoning_profile`	`standard \| deep`	deep is slower and costs more.

Response

Field	Type	Meaning
`jobId / status`	`uuid / queued \| processing`	The first claim's job handle to poll.
`claims[]`	`{claim, question, jobId, status} \| {claim, question, error}`	Every claim run, each its own job with its own result on /status.

Try it

POST /api/public/truth/submit

Response

// the response will appear here

Call it from your code

from the reader's browser (session cookie)

curl -sS -X POST 'https://<your-shepherd-host>/api/public/truth/submit' \
  -H "Cookie: <session cookie of the signed-in reader; browser only, no service bearer>" \
  -H 'content-type: application/json' \
  --data-binary @- <<'JSON'
{
  "content": "God won't give you more than you can handle.",
  "context": "Said to comfort a friend whose dad just died."
}
JSON

Browser only, with credentials: "same-origin". A service bearer gets 403 browser_session_required.

Versions

Every response from every Shepherd engine — a result, an estimate, a refusal or an error — says which versions produced it, under the same three headers on every engine. A JSON object body carries the same three as `versions: { api, engine, engines }`.

Response headers

Field	Type	Meaning
`X-API-Version`required	`v<N>`	The API version: the request/response contract, the same for every engine ("v1"). It changes only for a breaking change to that contract.
`X-Engine-Version`required	`<engine>-engine/<line>-v<N>`	The responding engine's version ("truth-engine/research-v32", "verse-engine/card-v1"). The number goes up whenever the engine's behaviour changes: its prompts, its models, its retrieval or its rules.
`X-Engine-Versions`required	`JSON object`	The versions of the engines this response's own work called, by engine name ({"lifeline": "lifeline-engine/match-v1", "truth": "truth-engine/research-v32"} from a FORMATION study; {} when it called none — a refusal before any work, or a result served from a recent identical run).

· Read as a label, a version is the engine's name, its number and the API's: "VERSE v3 · API v1". The example results on every engine page are labelled this way, with the versions each was captured under.
· A result stored for later (a TRUTH read, a FORMATION study's reviews) keeps the version that produced it; the headers name the versions answering now.
· TRUTH's assessment endpoints return the frozen assessment contract, which carries its own provenance.truth_engine_version, so there the versions are in the headers only. The open contract endpoint runs the deterministic stub and answers as truth-engine/stub-v0.