SWIFT ANALYZER / RULE CATALOG
138 rules about whole-program Swift behavior, from state that becomes stale across await to repeated collection work and sensitive data reaching logs.
Rules: 138 · conditions: 141
[email protected]Unused declaration or declaration group. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Production code used only by tests. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Forbidden dependency or layer bypass. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Excessive API visibility. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Unused product dependency. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Undeclared direct provider. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Disallowed component cycle. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Implementation type exposed through an API. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Incompatible registration or entry point. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Required resource missing from a shipping variant. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Parameter unused across its dispatch family. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Property is written but never observed. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Redundant protocol or conformance. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Semantically redundant import. It examines repository-wide reachability, dependencies, resources, and API boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Call crosses incompatible isolation. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]MainActor entry without a proven context. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Required context lost through detachment. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Callback runs in an incompatible scheduler context. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Invalid isolation assumption. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Resource used outside its execution domain. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Thread-affine value survives a suspension. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Required task-local value or executor preference is lost. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Blocking call on a critical executor. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Expensive computation on a critical executor. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Removable scheduling boundary. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Cost amplified by actor crossings. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Independent work is unintentionally serialized. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]Custom SerialExecutor violates serialization. It examines isolation, executor, scheduler, and execution-domain relationships. The result remains unknown when the required evidence is incomplete.
[email protected]State check becomes stale across await. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Broken invariant is observable during suspension. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Stale result overwrites a newer generation. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Operation requiring single-flight starts more than once. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Code depends on an unspecified actor-task order. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Continuation can finish without resume. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Continuation can resume more than once. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Cancellation is lost between registration and start. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Forbidden commit after cancellation or rejection. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Work does not honor its cancellation contract. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Timeout or first result does not finish its scope. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Task outlives its required lifecycle scope. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Required task result or error is lost. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Guard is held across a disallowed await. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Wait cycle blocks required producers. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Synchronous waits exhaust the cooperative pool. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Priority inversion through opaque waiting. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Shared-memory race across an unsafe or legacy boundary. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Synchronous callback re-enters an unfinished protocol. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Custom executor loses or repeats a job. It examines task ordering, suspension, cancellation, progress, and async obligations. The result remains unknown when the required evidence is incomplete.
[email protected]Borrowed pointer or view escapes its lifetime. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Use after free. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Repeated or incompatible deallocation. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Unowned access after owner deallocation. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Unmanaged ownership imbalance. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Strong ownership cycle without a valid termination. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Required weak consumer is released too early. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Registration outlives its allowed scope. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Cleanup obligation is not discharged. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Closed or transferred resource is still used. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Read from an uninitialized buffer region. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Pointer or count exceeds its valid extent. It examines ownership, resource state, and lifetime obligations across program paths. The result remains unknown when the required evidence is incomplete.
[email protected]Repeated linear traversal. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Repeated invariant sorting. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Repeated consumed materialization. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Array used as a growing shifting FIFO. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Full sort used for an extremum or top-k result. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Expensive projection repeated by a comparator. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Eager materialization before a short consumer. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Repeated physical copy-on-write detach. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Get-copy-modify-set amplifies cost. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Capacity policy violates the amortized bound. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Repeated element-by-element bridging. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Repeated expensive hashing or equality work. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Small slice retains large storage. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Fine-grained tasks amplify scheduling work. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Fanout retains excessive live data. It examines repeated work, allocation, traversal, retention, and scheduling cost. The result remains unknown when the required evidence is incomplete.
[email protected]Sensitive data reaches logs. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]SQL injection. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Shell injection. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Format-string injection. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Path traversal outside the extraction root. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]JavaScript or HTML context injection. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Unprivileged input controls a privileged URL sink. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Predicate injection. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Regular-expression injection. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Algorithmic denial of service through a regular expression. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]External-entity expansion from untrusted XML. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]TLS trust evaluation is effectively disabled. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Cryptography does not provide the claimed property. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Keyed nonce or IV is reused in an unsafe mode. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Unsafe password-derivation contract. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Predictable randomness in a security-sensitive value. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Untrusted deserialization crosses the allowed-type boundary. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Privileged effect bypasses required authorization. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Security validator or sanitizer is bypassed. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Embedded secret reaches a security-sensitive use. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]WebView receives forbidden capabilities. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Cryptographic key has insufficient strength. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Dependency is fetched over an untrusted channel. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Non-reproducible dependency resolution violates policy. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Debug capability is exposed in a shipping product. It examines trust boundaries, sensitive flows, validation, and security-sensitive effects. The result remains unknown when the required evidence is incomplete.
[email protected]Equality and hashing disagree. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Process-local hash is used as a stable identity. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Incompatible length or index units. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Invalidated collection index is used. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Key changes equality or hash while in a collection. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Ordering laws are violated. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Unstable ordering reaches canonical output. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Size or count exceeds its numeric range. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Non-finite value reaches a finite-only consumer. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Interprocedural path reaches an invalid force operation. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Single-pass sequence is consumed more than once. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Copy violates the state-independence contract. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Producer and consumer canonicalize identity differently. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Written value is lost before observation. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Predicate does not distinguish valid states. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Conditional-result branches are semantically identical. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Control-flow region is unreachable. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Computed result is discarded without a useful effect. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Comparison violates its numeric precision contract. It examines value, collection, identity, ordering, and control-flow contracts. The result remains unknown when the required evidence is incomplete.
[email protected]Buffer or accumulator lacks a required bound. It examines buffering, delivery, progress, retry, and transaction boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Drop policy violates required event delivery. It examines buffering, delivery, progress, retry, and transaction boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Concurrent next calls on a single-consumer iterator. It examines buffering, delivery, progress, retry, and transaction boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Stream is not finished on a required terminal path. It examines buffering, delivery, progress, retry, and transaction boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Retry or refresh forms a loop without progress or budget. It examines buffering, delivery, progress, retry, and transaction boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Retry duplicates a non-idempotent effect. It examines buffering, delivery, progress, retry, and transaction boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Related changes escape a required transaction. It examines buffering, delivery, progress, retry, and transaction boundaries. The result remains unknown when the required evidence is incomplete.
[email protected]Element identity is unstable across updates. It examines SwiftUI identity, update, and state-lifecycle behavior. The result remains unknown when the required evidence is incomplete.
[email protected]Distinct elements share one identity. It examines SwiftUI identity, update, and state-lifecycle behavior. The result remains unknown when the required evidence is incomplete.
[email protected]View computation causes a state-update cycle. It examines SwiftUI identity, update, and state-lifecycle behavior. The result remains unknown when the required evidence is incomplete.
[email protected]One-time state does not update with required input. It examines SwiftUI identity, update, and state-lifecycle behavior. The result remains unknown when the required evidence is incomplete.
[email protected]Source API change breaks an admitted consumer. It examines source, ABI, dependency, resource, and cache compatibility across changes. The result remains unknown when the required evidence is incomplete.
[email protected]ABI change breaks binary compatibility. It examines source, ABI, dependency, resource, and cache compatibility across changes. The result remains unknown when the required evidence is incomplete.
[email protected]Existing signature gains incompatible behavior. It examines source, ABI, dependency, resource, and cache compatibility across changes. The result remains unknown when the required evidence is incomplete.
[email protected]Consumer silently binds to a different symbol. It examines source, ABI, dependency, resource, and cache compatibility across changes. The result remains unknown when the required evidence is incomplete.
[email protected]Version range admits an incompatible provider. It examines source, ABI, dependency, resource, and cache compatibility across changes. The result remains unknown when the required evidence is incomplete.
[email protected]Change bypasses required consumer review. It examines source, ABI, dependency, resource, and cache compatibility across changes. The result remains unknown when the required evidence is incomplete.
[email protected]Comparable resource bound regresses. It examines source, ABI, dependency, resource, and cache compatibility across changes. The result remains unknown when the required evidence is incomplete.
[email protected]Cache key or invalidation omits a result dependency. It examines source, ABI, dependency, resource, and cache compatibility across changes. The result remains unknown when the required evidence is incomplete.
One RuleID can group several conditions. Result kind does not define severity, and first-result strategy does not imply readiness.