Struct NamedAgentRecordV1 

pub struct NamedAgentRecordV1 {
    pub instance_id: String,
    pub instance_name: String,
    pub definition_id: String,
    pub identity_id: Option<String>,
    pub memory_id: Option<String>,
    pub session_id: Option<String>,
    pub working_dir: String,
    pub source_agents_base: Option<String>,
    pub created_at_ms: i64,
    pub last_launched_at_ms: i64,
    pub created_by_version: String,
    pub last_launched_by_version: String,
}

v1 payload. Add new optional fields here under #[serde(default)] and bump MAX_SUPPORTED_SCHEMA — old readers will skip the new version, new readers fill defaults for old files.

Fields

instance_id: String

instance_name: String

definition_id: String

identity_id: Option<String>

FK to db_identity_bundles.id. None = unbound (= ambient creds).

memory_id: Option<String>

FK to db_memory_bundles.id. None = unbound (= vanilla CLI).

session_id: Option<String>

Provider CLI session id for --resume (e.g. a Claude Code session uuid). None = no session wired yet (fresh stub / legacy record). Added in schema v2 so a global/migrated record can resume the agent’s conversation across channels without a current-channel SQLite join. Old (v1) files deserialize this as None.

working_dir: String

Path relative to [source_agents_base] (or, for legacy records without one, the reader’s current channel agents dir) — never absolute. Keeps the record portable across machines where the home dir differs.

source_agents_base: Option<String>

Absolute path of the agents dir that [working_dir] is relative to — i.e. the channel/dev instance the agent actually lives in (channels/<ch>/agents or <dev-instance>/agents). Added in schema v3 (cross-channel persistence P0.4): the registry is global, so a row surfaced in a DIFFERENT channel must reconstruct its absolute working_directory against the SOURCE base, not the reader’s current channel. None for legacy (v1/v2) records — the reader then falls back to its current channel agents dir, matching pre-P0.4 behavior. Old binaries deserialize this as None.

created_at_ms: i64

last_launched_at_ms: i64

created_by_version: String

last_launched_by_version: String

Implementations

impl NamedAgentRecordV1

pub fn min_schema_version(&self) -> u32

Lowest envelope schema that can faithfully represent this payload. Climbs only when a higher-version-only field is populated, so an older binary keeps reading records that don’t use any newer feature — only records that actually need the newer schema are hidden from it. Writers should stamp the record with this rather than always using MAX_SUPPORTED_SCHEMA.

• v3 when source_agents_base is set (cross-channel reconstruction)
• v2 when session_id is set (cross-channel resume)
• v1 otherwise

Trait Implementations

impl Clone for NamedAgentRecordV1

fn clone(&self) -> NamedAgentRecordV1

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for NamedAgentRecordV1

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for NamedAgentRecordV1

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl PartialEq for NamedAgentRecordV1

fn eq(&self, other: &NamedAgentRecordV1) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for NamedAgentRecordV1

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Eq for NamedAgentRecordV1

impl StructuralPartialEq for NamedAgentRecordV1