pub struct Chunk {Show 23 fields
pub id: String,
pub project: String,
pub agent: String,
pub date: String,
pub session_id: String,
pub cwd: Option<String>,
pub kind: Kind,
pub frame_kind: Option<FrameKind>,
pub run_id: Option<String>,
pub prompt_id: Option<String>,
pub agent_model: Option<String>,
pub started_at: Option<String>,
pub completed_at: Option<String>,
pub token_usage: Option<u64>,
pub findings_count: Option<u32>,
pub workflow_phase: Option<String>,
pub mode: Option<String>,
pub skill_code: Option<String>,
pub framework_version: Option<String>,
pub msg_range: (usize, usize),
pub text: String,
pub token_estimate: usize,
pub highlights: Vec<String>,
}Expand description
A single chunk ready for vector indexing.
Fields§
§id: StringUnique ID: {project}_{agent}_{date}_{seq:03}
project: String§agent: String§date: StringDate string (YYYY-MM-DD)
session_id: StringSession ID from first message in chunk
cwd: Option<String>Working directory from the first message in the chunk window
kind: KindClassified kind for this chunk’s content
frame_kind: Option<FrameKind>Stable stream/channel classification for the chunk contents.
run_id: Option<String>Optional correlation ID for the originating run
prompt_id: Option<String>Optional prompt or task identity for the originating run
agent_model: Option<String>Optional agent model reported by the source frontmatter
started_at: Option<String>Optional run start timestamp reported by the source frontmatter
completed_at: Option<String>Optional run completion timestamp reported by the source frontmatter
token_usage: Option<u64>Optional token usage reported by the source frontmatter
findings_count: Option<u32>Optional findings count reported by the source frontmatter
workflow_phase: Option<String>Optional workflow phase reported by the source frontmatter
mode: Option<String>Optional routing mode reported by the source frontmatter
skill_code: Option<String>Optional framework skill code reported by the source frontmatter
framework_version: Option<String>Optional steering schema/framework version reported by the source frontmatter
msg_range: (usize, usize)Index range in original day’s entries (start, end exclusive)
text: StringFormatted chunk text with header
token_estimate: usizeEstimated token count (~chars/4)
highlights: Vec<String>Decision/plan highlights extracted from the chunk