FACTUAL ACTIVITY RECORD · An activity record based on work that took place
Inventorying activity evidence at a fixed cutoff
Establishing record-level traceability and candidate clearance without confusing retrieval with publishability
Purpose
The evidence snapshot supporting the first activity candidates ended at a number boundary, leaving later and updated records and state changes unavailable for comparison at one point in time.
The work needed a fixed scope and observation time, a fresh source inventory that kept completed, not adopted, in-progress, future, and unverified states distinct, and traceability from candidate facts back to their originals.
Implementation
The scope was limited to one source area, and every number through an exclusive cutoff was classified as a work or change record. The inventory covered 61 work records, 41 change records, and 325 cutoff-valid comments, checking missing numbers, duplicates, and post-cutoff inclusion at record level.
A monolithic query exceeded the node limit and was rejected, so population, work-record, and change-record retrieval were divided into bounded queries. Every connection had to complete pagination; partial responses, remaining pages, API failure, and count mismatches stopped the inventory from claiming full coverage.
A record-level source manifest was created separately from the inventory, retaining source type, state, timestamps, comment, conversation, review and changed-file evidence, classification, exclusion reasons, and current-artifact mapping. Values at the fixed observation were separated from later current references so mutable values could not rewrite the snapshot.
- Treat
confirmed,excluded, andneeds-reviewas retrieval states, separate from facts, editorial decisions, clearance, and publication. - Cross-check completed, not-adopted, and unstarted or in-progress work against bodies, comments, related changes, and current artifacts.
- Mark evidence from another source area as
external-unverifiedrather than treating parent-side records as direct verification. - Do not copy private data, credentials, personal data, or unknowns into public data, and do not reuse internal manifest URLs, object IDs, or paths directly as public evidence.
Narrative checks separated retrieval state from content class and applied interpretation boundaries without changing candidate mappings, calls to action, or public copy. Legal and reputation assessment recorded candidate-level unknowns for privacy, rights, claim basis, data processing, and human review; any unresolved field kept clearance at HOLD.
Because copying live workflow status into a fixed artifact would make it stale, the inventory and manifest retain observed evidence only, while current progress is obtained from external authoritative sources. Story-specific acceptance was also separated from downstream publication gates.
The source manifest encoded the fixed-observation rule and retrieval stop conditions as follows. The public account records only snapshot-wide invariants, not identifiers for individual records.
Rendering diagram…
What was confirmed
Verification results
All 102 numbers in the fixed range were classified without duplication into 61 work records and 41 change records, with 325 cutoff-valid comments and related evidence checked through complete pagination.
Completed, not-adopted, future, and in-progress states remained distinct, while other source areas stayed external-unverified and incomplete retrieval stayed fail-closed as needs-review.
Publication clearance remained HOLD for every candidate, with zero READY candidates, so retrieved evidence did not imply publishability.
At the cutoff, the inventory, record-level manifest, and references to the content model had been applied to the target. Candidate public copy, human legal, compliance, and locale review, and publication approval were incomplete.
Basis for completion
The inventory was complete when the fixed cutoff, full number coverage, pagination, record-level traceability, fixed observation, separation of retrieval and publication decisions, and candidate clearance were recomputed and applied consistently across four canonical artifacts.