A trust channel is the identity and verification context that lets a receiver judge who sent an exchange and which credentials or signatures matter.
Identity
Which actor, principal, service, or organization is represented.
Credential
Which credential, signature, DID, VC, token, or authority should be checked.
Boundary
Which trust assumption applies inside the runtime versus across a public handoff boundary.
In an LLM Wiki
Trust-channel notes belong on pages that describe cross-system handoffs, signed artifacts, source provenance, contributor actions, or automation with tool permissions. A useful note says who acted, what credential or authority was used, what was verified, and what remains an assumption.
Review questions
- Can a later reader identify the actor or system that produced the exchange?
- Is the credential, token, key, signature, or authority named without exposing secrets?
- Does the page distinguish internal trust from public handoff trust?
- Is the trust claim linked to provenance, integrity, and delivery evidence?