This map answers the visitor question: “Show me which tested data supports this case.” It connects the public case narratives to the participant-free synthetic package and states where the connection is only partial or has not yet been instantiated.

The machine-readable version is case_data_map.csv.

Case 001 — Capability Blockage

Mapping status: Partial alignment

The current proxy is OSI-SYN-SRC-002 → OSI-SYN-EVD-002 → OSI-SYN-OBS-002. The fixture records decisions routed through a single operational role. That is relevant to a blockage hypothesis, but it does not encode stable capability, workload, utilization decline, or causality.

Case 002 — False Capability Signal

Mapping status: Bounded match

The tested path is OSI-SYN-SRC-003 → OSI-SYN-EVD-003 → OSI-SYN-OBS-003. The evidence explicitly records delivery continuity while the team relies on undocumented role knowledge. This supports the bounded output-versus-resilience distinction, not a claim about organizational health or future failure.

Case 003 — Misattribution of Failure

Mapping status: Not yet instantiated

The current synthetic package has no evidence row encoding conflicting reporting paths, unclear authority, or individual attribution. The case is therefore a public research scenario, not a claim that the current sandbox run has tested. This distinction is intentional and prevents the graph mechanics from being presented as stronger evidence than they are.

The next data increment should add a dedicated synthetic record and a test that asserts the case-specific mapping before this case is called demonstrated.

What the tests prove today

The assurance suite proves the package’s file contracts, provenance links, synthetic-only identities, review gates, confidence bounds, and negative boundaries. It does not prove that a narrative case is semantically reproduced by a fixture unless this map says so.