pub struct TrustConfig {
pub default_level: SkillTrustLevel,
pub local_level: SkillTrustLevel,
pub hash_mismatch_level: SkillTrustLevel,
pub bundled_level: SkillTrustLevel,
pub scan_on_load: bool,
pub scanner: ScannerConfig,
}Expand description
Skill trust policy configuration, nested under [skills.trust] in TOML.
Controls how trust levels are assigned to skills at load time based on their origin (local filesystem vs network) and integrity (hash verification result).
§Example (TOML)
[skills.trust]
default_level = "quarantined"
local_level = "trusted"
scan_on_load = trueFields§
§default_level: SkillTrustLevelTrust level assigned to skills from unknown or remote origins. Default: quarantined.
local_level: SkillTrustLevelTrust level assigned to skills found on the local filesystem. Default: trusted.
hash_mismatch_level: SkillTrustLevelTrust level assigned when a skill’s content hash does not match the stored hash.
Default: quarantined.
bundled_level: SkillTrustLevelTrust level assigned to bundled (built-in) skills shipped with the binary. Default: trusted.
scan_on_load: boolScan skill body content for injection patterns at load time.
When true, SkillRegistry::scan_loaded() is called at agent startup.
This is advisory only — scan results are logged as warnings and do not
automatically change trust levels or block tool calls.
Defaults to true (secure by default).
scanner: ScannerConfigFine-grained scanner controls (injection patterns, capability escalation).
Trait Implementations§
Source§impl Clone for TrustConfig
impl Clone for TrustConfig
Source§fn clone(&self) -> TrustConfig
fn clone(&self) -> TrustConfig
1.0.0 (const: unstable) · Source§fn clone_from(&mut self, source: &Self)
fn clone_from(&mut self, source: &Self)
source. Read more