go-security

name	go-security
description	Security: Use when trust boundaries, identity, authorization, tenant isolation, browser/token controls, secrets, abuse, injection, or SSRF needs a decision, or when changed Go crosses those boundaries. Own security policy and conformance review; Skip when the primary concern is non-security API, data, reliability, or implementation placement.

Go Security

Load the shared specialist contract. Reconstruct every reachable attacker path from affected trust-boundary surfaces in routes, handlers, identities, objects, stores, outbound calls, secrets, and abuse controls. Follow each path through identity, authority, asset, enforcement point, attacker action, and observable failure; silence is never evidence that a path is closed.

Choose The Branch

• Decision — select when security policy is absent or changing. Load the decision selector for one result-changing threat. Complete when shared Decision dispositions cover every reachable path with fail-closed behavior and focused negative proof.
• Review — select when changed code must conform to accepted security policy. Load the review selector for the concrete attacker path. Follow every reachable path into the shared finding envelope, naming any outside boundary or proof blocker; no finding requires focused negative proof. Missing policy returns to the named Security Decision owner.

Hand non-security contract semantics to go-api-contract, data authority to go-data-architecture, and placement to go-implementation-ownership.