Non-Official Cover (NOC)

Non-official cover (NOC) is the intelligence-tradecraft practice of operating under an identity with no overt government affiliation: a cover job, business, or profession — the legend — that explains the operative's presence, income, and movements while masking compartmented work. Unlike official cover (a diplomatic or other government posting), a NOC legend must survive scrutiny entirely on its civilian paper trail: licenses, corporate filings, employment history, publications.

How this vault uses the term

Several whistleblower-type figures in this vault claim NOC-style identities — a public/official job they say masked compartmented UAP work (canonical example: Jake Barber). This note anchors the vault's cover legend analysis convention:

• The covert role is unverifiable by design; the pipeline never tries to settle it.
• The cover side lives in the public record — pilot certificates, patents, corporate registries, contracts — and can be checked. That is where the analysis does its work.
• The deliverable is a per-element verification table of the claimed legend, never a conclusion about whether the subject "really was" NOC.

Verdict vocabulary (one per claimed legend element):

Three epistemic rules:

1. Signals cut both ways. A legend that checks out proves the legend exists, not that it was a cover. A legend that fails is a strike against the whole account — the same rubric that could support a real NOC catches a fabricator using "it was classified" as cover for having no evidence. Both outcomes are findings.
2. No conclusion fields. No suspected_noc-style boolean ever appears in frontmatter, reports, or the graph. The verification table is the artifact; inference stays with the reader. A boolean would launder our inference into the graph as fact.
3. Only verify what the subject publicly claimed about themselves. The lane checks self-made public claims against public records. It does not compile private information about private individuals, and it does not extend to people who never claimed a cover.

Discrepancy-signal rubric

Observations worth flagging in an analysis section — each is a signal, not a verdict, and each cuts both ways (legend-craft or embellishment):

1. Overloaded résumé — more simultaneous roles than one person plausibly holds.
2. Clearance/role mismatch — the stated job doesn't warrant the documented or claimed access ("airplane mechanic with NATO Top Secret").
3. Employer opacity — the cover employer is vague ("a helicopter company"), a shell with no registry/contract footprint, or a known SAP-adjacent contractor.
4. Skill overshoot — certifications or tradecraft exceeding the stated role (a mechanic doing red-team threat assessments).
5. Geographic/timeline mismatch — documented locations versus claimed operational locations and dates.
6. Footprint asymmetry — the cover side is too clean (no coworkers, press, or filings for a claimed company president), or the subject's records are sealed where peers' aren't.

The cluster this note enables

Every Person note carrying a ### Cover Legend Analysis section wikilinks this note, making it a queryable cluster node: everyone with an alleged legend is one hop away. The follow-on question the cluster enables: do alleged covers converge on shared shell employers or contractors? Shared cover infrastructure across independent claimants would be a genuine finding.

The research pipeline runs this lane via the cover_legend_analysis gap type in packs/uap/pack.yaml; the section convention is documented in Graph Schema.