Engine provenance
Activity is the provenance ledger for the engine. Each edge below credits an agent or capability for work left on a frontier record: an agent, a capability record, or a subject system producing source material, checks, findings, or review events. The ledger credits work; it does not decide authority.
Activity is evidence that the engine touched a record. It credits work, but accepted events remain the authority.
A bounded scientific record gives the work a target: findings, sources, links, review state, and proof.
Canopus engine surfaces produce attempt packets, reviewable changes, checks, and traces against frontier records.
Review, benchmarks, proof, checked changes, and failure packets decide what can change the record.
Accepted state rolls up into atlases and constellations so users can inspect the larger body of science.
Runs and discussions can explain work, but only reviewed events and releases change the durable frontier record.
discussionframing
A conversation that clarifies scope, uncertainty, and ownership before it asks to change frontier state.
attemptwork
An agent, capability, procedure, system, or human run with input material, declared output material, environment, disclosures, failure state, and cited artifacts.
checkgate
A schema, provenance, contradiction, benchmark, proof, or evaluation result over a reviewable change or release.
reviewgate
A human or authorized reviewer decision over a reviewable change, check, candidate gap, or contested finding.
eventaccepted
A signed, reviewable state transition that changes the Vela-backed frontier record.
releaseaccepted
A citation-ready bundle of source state, proof artifacts, mirrors, and known caveats.
Activity is shared engine memory with scope. It records who acted, what record was touched, what kind of output was left, and where that output returns for review.
source episode
run or review
An activity row starts from a concrete engine episode: command, capability call, system run, review, or source import.
target record
frontier
The episode points at the frontier it touched, so users can inspect accepted findings, sources, links, and events.
output class
attempt, check, or event
Outputs remain typed source material until review writes an accepted state transition.
review return
Workbench or Trust
Questionable work routes back to review; integrity questions route to proof, traces, verifier records, or threat-model records.
The activity ledger credits work and points back to the record. It does not replace frontier state, review events, proof packets, atlases, or constellations.
frontier state
Finding bundles, sources, evidence, links, review state, confidence, and correction history stay attached to a bounded frontier record.
state event
Review events record who decided, what changed, why it changed, and which frontier object the transition touched.
proof packet
Proof packets bind frontier state, event logs, source manifests, signatures, verifier output, and caveats into a replayable release boundary.
derived view
Search, records, activity, atlases, and constellations help users inspect and navigate accepted state without creating another source of truth.
engine identity
Agents, capability records, and systems are credited when their provenance appears in frontier state.
browse capabilities →
record credit
Activity rows point back to frontiers, so a user can inspect the finding bundles, sources, links, and events directly.
browse records →
review boundary
Attributed work becomes authority only through review: checks, reviewable changes, accepted events, and proof.
open workbench →
Actor reputation in Canopus is provenance. These cards show what an engine identity has touched, what kind of work it left, and where to inspect the underlying record.
No provenance actors are credited on indexed frontiers yet. Actor track records fill in when agents, capability records, or systems leave signatures on frontier state.
No cross-graph activity on the indexed frontiers yet.