Receipt reference · the artefact

Every line a run prints, and what it binds.

A run ends in one receipt: one line per field, printed to standard output, and nothing that summarises it away. This page is the reference for those lines — what each one binds, how to read the ones that are easy to misread, and three worked readings from real runs.

The field reference Fields that mislead Worked readings

live:falsesimulation-onlyabstain is a resultrejection settles nothing

The artefact

One run, one receipt, one line per field.

The entry prints the receipt to standard output and exits zero. A refusal prints a named error instead and exits non-zero, so an absent receipt is itself a readable result.

receipt shape

entry=kalkas
domain=<sport|quant|financial>
scenario=<act, or a named boundary>
pit_repository=<sealed source authority>
pit_commit=<source revision>
pit_protocol=<source row protocol>
pit_corpus_digest=<content seal over the admitted corpus>
promotion=<accepted|rejected|none>
rejection=<named reason|none>
decision=<act|abstain|none>
decision_run_id=<decision identity|none>
abstention=<named reason|none>
effect=<simulation|closed_read_only|none>
effect_receipt=<adapter receipt class|none>
execution_id=<effect identity|none>
settled=<true|false>
evaluated=<true|false>
outcome_bound=<true|false>
learning_bound=<true|false>
replay_verified=<true|false>
north_star=<verified|none>
weekly_verified=<verified runs in this local history, last seven days>
retry_effects=<adapter executions this run recorded; readbacks are not counted>
live=false
ordinary=none private engine, no public web surface
preview=none local product entry, not a hosted runtime

Field shape, not a captured run: the entry prints one line per field, with the sealed identities, digests, and counts in place of the placeholders.

ScopeRead it as a statement about one run

Every line describes what happened on this run: which corpus was read, which gate answered, which decision was emitted, and whether the outcome and the learning receipt bound to it.

LimitWhat not to read into it

No line is a performance number. promotion, decision, settled, and replay_verified describe what happened here, not what any run will earn. The count on the weekly line is a local count of verified completed runs, and the effect lines describe a simulation.

Field reference

Every field, in the order it prints.

The table below is the whole surface: no line is omitted and no line is upgraded. Where a value is a closed set, the set is written out.

Every receipt field, what it binds, and how to read it
Field	What it binds	How to read it
entry	The entry point that produced the receipt.	One line, one run. Every receipt starts here.
domain	The domain module the run bound to.	Names which sealed corpus supplied the evidence.
scenario	The boundary the run exercised.	act for the admitted path; otherwise the named boundary you asked for.
pit_repository	The sealed source authority that produced the corpus.	Provenance, printed in full so it can be checked rather than assumed.
pit_commit	The revision of that source the corpus was sealed at.	Two receipts naming the same revision read the same sealed state.
pit_protocol	The row protocol the corpus follows.	Says which field law the evidence was admitted under.
pit_corpus_digest	The content seal over the admitted point-in-time corpus.	The same seal in two receipts means the same corpus was read.
promotion	Whether the candidate cleared the promotion gate.	accepted, rejected, or none when the gate was never reached. A rejection stops the run before any decision.
rejection	The named reason a promotion did not happen.	One of dataset_not_oos, coverage_below_floor, loss_above_ceiling, or improvement_below_floor; none on the admitted path.
decision	The typed outcome of the run.	act, abstain, or none. none means no decision was emitted, which is what a rejection produces.
decision_run_id	The identity of the decision.	Settlement binds to this identity. If it is none, there is nothing to settle.
abstention	The named reason the run chose not to act.	missing_mandate, invalid_mandate, expired_mandate, prohibited_effect, low_confidence, exposure_limit, ruin_policy, or domain_declined; none when the run acted.
effect	The effect class the run used.	simulation, closed_read_only, or none. There is no live variant: nothing external is contacted.
effect_receipt	Which adapter produced the effect receipt.	none, sport_simulation, finance_simulation, or vh_closed_read_only. The class cannot represent a live order, trade, bet, or transfer.
execution_id	The identity of the effect.	Bound to the run, the decision, and the idempotency key that requested it; none when no effect was admitted.
settled	Whether the outcome resolved and bound to this run.	true for acted and abstained runs that reached settlement; false when a rejection left nothing to settle.
evaluated	Whether evaluation ran.	Rejections evaluate too: the gate result is recorded even when nothing settles.
outcome_bound	Whether the resolved outcome bound to this exact run.	Read with settled. It holds for abstentions as well as acts, and only where an execution exists is the effect identity part of the binding.
learning_bound	Whether the learning receipt bound to this run.	Holds for abstentions as well as acts. A rejection emits no learning receipt at all.
replay_verified	Whether replay reproduced the decision from the same inputs.	true means the recorded decision is the one those inputs produce, and replay called no adapter.
north_star	The verification status of this run.	verified or none. none means the run did not complete a decision to verify.
weekly_verified	The count of verified completed runs in the local history over the last seven days.	A local count over the history file you pointed at. Rejected candidates leave it unchanged.
retry_effects	How many adapter executions this run recorded.	Reading the original receipt back is not a new execution, so the count stays put.
live	The exposure encoding of the run.	Always false today. The shipped adapter has no live path.
ordinary, preview	The product stating its own exposure in words.	No public web surface, and a local entry point rather than a hosted runtime.

Read twice

Nine fields that are easy to misread.

These lines are unambiguous once you know what they are not. Each row below says what the field records and what a reader might wrongly take from it.

What the field records, and what it does not
Field	What it records	What it is not
settled	The outcome resolved and bound to this run — for an act and for an abstention alike.	Not a measure of whether the decision was right, and not a promise of a gain. A run that abstained settled just as completely as one that acted.
evaluated	The gate result was recorded, including on the path where the candidate was rejected.	Not a statement that anything settled or was learned. A rejection evaluates and still ends the run before planning.
outcome_bound	The resolved outcome carries this decision run identity (and, when an execution exists, that effect identity too).	Not evidence that an effect happened: an abstention binds its outcome with no execution at all.
learning_bound	A learning receipt exists and names this run.	Not a claim that the lesson was favourable, and not a requirement for an execution: learning does not need one.
replay_verified	Replay reproduced the decision from the same inputs and the same model version, with zero adapter calls.	Not a re-run of the effect and not a second opinion: it is the same decision, recomputed.
north_star	The verification status of this one run against the seven conditions.	Not a weekly total and not a performance figure. Weekly verification is the count on the line below it.
weekly_verified	The count of verified completed runs in the window, in the history file this run used.	Not a production count, and not a tally of calls, orders, or bets. Rejected candidates never move it, and a fresh history file starts it again.
retry_effects	The adapter executions this run actually recorded.	Not a retry counter: reading the original receipt back does not add to it, and a conflicting retry is refused rather than counted.
live	The published encoding of the run’s exposure: false.	Not a switch, not a staging value, and not something a flag can change. There is no live variant in the shipped adapter.

The seven conditions

Where to check each condition in the receipt.

north_star is the run’s verdict on all seven. The table pairs each condition with the lines that let you check it yourself, and says plainly where the evidence is sealed rather than printed.

Each condition, the receipt lines that carry it, and what is sealed rather than printed
Condition	Receipt lines	What to know
Evidence carries observation time, availability time, and provenance.	pit_repository, pit_commit, pit_protocol, pit_corpus_digest	The observation time, the availability time, and the source identity travel inside the decision identity; a fact that was not available at the decision clock stops the run before a receipt exists.
Replay over the same inputs and model version is deterministic.	replay_verified	Replay recomputes the decision from the same admitted inputs and model version, and a mismatch stops the run rather than being recorded as verified.
The belief exposes uncertainty and source lineage.	decision_run_id, north_star	The belief value, its uncertainty, and its model and dataset lineage are sealed into the decision identity; a belief without usable uncertainty is refused before it reaches a receipt.
The mandate admits or refuses without an override.	promotion, decision, abstention	A missing, expired, or invalid mandate, a prohibited effect, or a breached bound records an abstention reason instead of an effect request.
The run ends in a typed act or abstention receipt.	decision, decision_run_id	Exactly one typed outcome per run, replayable from the same inputs; a rejection produces none.
The outcome resolves and binds to that exact run.	settled, outcome_bound, execution_id	The outcome binds market and run identity together, with an effect identity only when an execution occurred.
Settlement emits evaluation and learning receipts.	evaluated, learning_bound, north_star	Evaluation and learning receipts are emitted for acted and abstained runs; the execution receipt is optional and the learning receipt is not.

Worked readings

An act, an abstention, and a rejection.

Three runs, read line by line. Each block below is the receipt shape with the lines that matter filled in and the identities left as placeholders — the values that settle, bind, and count.

01
One act run
act
Every gate passed, the act was admitted, and the effect adapter recorded one simulation under the run’s idempotency key.
Read the effect lines with the settlement lines. The effect is a simulation with its own identity bound to the run and the key; the run settles, learns, and replays, so it counts once in the local verified window. A retry of the same request reads that receipt back rather than recording a second execution, which is why retry_effects stays where it is.
kalkas run --domain sport
act · lines that matter
promotion=accepted rejection=none decision=act decision_run_id=<decision identity> abstention=none effect=simulation effect_receipt=sport_simulation execution_id=<effect identity> settled=true evaluated=true outcome_bound=true learning_bound=true replay_verified=true north_star=verified weekly_verified=<count, this run included> retry_effects=1
Receipt shape with the deciding lines filled in. Identities, digests, and counts stay as placeholders.
02
One abstention
abstain
The gate admitted the candidate, and the belief sat inside the zone where acting is not justified, so the run abstained with that reason recorded.
An abstention is a completed run: it settles, binds its outcome, emits its learning receipt, and stays in coverage. What it does not have is an effect. The abstention line names the reason, and the weekly count treats this run exactly as it treats an act.
kalkas run --domain sport --scenario abstain
abstain · lines that matter
promotion=accepted rejection=none decision=abstain decision_run_id=<decision identity> abstention=low_confidence effect=none effect_receipt=none execution_id=none settled=true evaluated=true outcome_bound=true learning_bound=true replay_verified=true north_star=verified weekly_verified=<count, this run included> retry_effects=0
Receipt shape with the deciding lines filled in. Identities, digests, and counts stay as placeholders.
03
One rejection
reject
The candidate failed a declared promotion criterion, so the run stopped before planning: nothing settles and no learning receipt is emitted.
This is the receipt to read carefully. evaluated is true because the gate result was recorded; settled, outcome_bound, learning_bound, and replay_verified are false because there is no decision to settle, no outcome to bind, and no learning receipt. The identity lines say none, the effect lines say none, and the local verified count does not move.
kalkas run --domain sport --scenario reject
reject · lines that matter
promotion=rejected rejection=<named reason> decision=none decision_run_id=none abstention=none effect=none effect_receipt=none execution_id=none settled=false evaluated=true outcome_bound=false learning_bound=false replay_verified=false north_star=none weekly_verified=<count, unchanged by this run> retry_effects=0
Receipt shape with the deciding lines filled in. Identities, digests, and counts stay as placeholders.

The quickstart runs these three boundaries end to end, and the refusal reference lists every boundary with the token that shows it.

Where the receipt is enforced.

The kernel reference explains the typed contracts each line is projected from. The verification method explains the replay, coverage, and settlement checks behind the word verified.

Kernel reference Verification method