51 crates, 753 tests green (up from 744 at v0.32.0). All 5 agents launched in parallel via substrate composed prompts. Zero file conflicts between scopes. ## A. kei-ledger v6 — 3 performance indexes + fork_transactional API - idx_agents_started_ts, idx_agents_status, idx_agents_fork_parent_id - fork_transactional<F>(): atomic fork row + caller side-effect closes kei-fork ↔ ledger transactional gap (Wave 15 known issue) - 30 tests (was 23) ## B. kei-prune dedupe — cluster-based retirement via kei-dna-index - dedupe_candidates() via kei_dna_index::cluster_by(Scope) - dedupe_strict() — intersection of scope+body clusters - apply_retirements() reuses mark_retired - CLI: kei-prune dedupe [--strict] [--dry-run] - 17 tests (was 9) ## C. kei-brain-view clusters — stats + cluster visualization - render_clusters(by: ClusterBy) — indented tree output - render_summary() — 6-line dashboard from kei-dna-index::stats - CLI: kei-brain-view clusters --by scope|body|role; kei-brain-view summary - 15 tests (was 8) ## D. kei-fork watch-hook — auto-collect on .DONE marker - watch_loop() — kei-watch subscription + auto kei-fork collect - hooks/fork-collect-on-done.sh — foreground daemon wrapper - dedupe via HashSet + agent_id validation - 21 tests (was 13) ## E. three-role pipeline (Writer → Auditor → Merger) - _roles/auditor.toml (pipeline.handoff=["merger"], claude-subagent-type=critic) - _roles/merger.toml (leaf, git-ops scoped, infra-implementer) - 5 new capability fragments: policy/git-ops-scope, scope/read-only, output/verdict, output/merge-result, verify/fork-audit - kei-spawn: pipeline_from_role + emit_pipeline_json + scaffold_downstream_tasks - kei-spawn: precedent::run_advisory (env-gated KEI_SPAWN_PRECEDENT_CHECK) - CLI: kei-spawn spawn --pipeline - 19 tests (was 10) Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
95 lines
3.5 KiB
Rust
95 lines
3.5 KiB
Rust
//! Integration tests for `render_summary`.
|
|
|
|
use kei_brain_view::render_summary;
|
|
use rusqlite::{params, Connection};
|
|
use tempfile::TempDir;
|
|
|
|
fn seed_db() -> (TempDir, Connection) {
|
|
let dir = tempfile::tempdir().unwrap();
|
|
let db = dir.path().join("ledger.sqlite");
|
|
let conn = Connection::open(&db).unwrap();
|
|
conn.execute_batch(
|
|
"CREATE TABLE agents (
|
|
id TEXT PRIMARY KEY,
|
|
branch TEXT NOT NULL,
|
|
parent_branch TEXT,
|
|
spec_sha TEXT NOT NULL,
|
|
status TEXT NOT NULL,
|
|
started_ts INTEGER NOT NULL,
|
|
finished_ts INTEGER,
|
|
summary TEXT,
|
|
worktree_path TEXT,
|
|
dna TEXT,
|
|
creator_id TEXT,
|
|
fork_parent_id TEXT
|
|
);",
|
|
)
|
|
.unwrap();
|
|
(dir, conn)
|
|
}
|
|
|
|
fn dna(role: &str, caps: &str, scope: &str, body: &str, nonce: &str) -> String {
|
|
format!("{role}::{caps}::{scope}::{body}-{nonce}")
|
|
}
|
|
|
|
fn insert(conn: &Connection, id: &str, status: &str, ts: i64, dna_str: Option<&str>) {
|
|
conn.execute(
|
|
"INSERT INTO agents
|
|
(id, branch, parent_branch, spec_sha, status, started_ts, dna)
|
|
VALUES (?1, ?2, NULL, 'deadbeef', ?3, ?4, ?5)",
|
|
params![id, format!("agent/{id}"), status, ts, dna_str],
|
|
)
|
|
.unwrap();
|
|
}
|
|
|
|
#[test]
|
|
fn render_summary_shows_all_fields() {
|
|
let (_d, conn) = seed_db();
|
|
let d1 = dna("edit", "NG-FW", "AAAAAAAA", "B1B1B1B1", "11111111");
|
|
let d2 = dna("edit", "NG-FW", "AAAAAAAA", "B2B2B2B2", "22222222");
|
|
let d3 = dna("edit", "NG-FW", "CCCCCCCC", "B3B3B3B3", "33333333");
|
|
insert(&conn, "a1", "running", 1, Some(&d1));
|
|
insert(&conn, "a2", "done", 2, Some(&d2));
|
|
insert(&conn, "a3", "failed", 3, Some(&d3));
|
|
|
|
let out = render_summary(&conn).unwrap();
|
|
assert!(out.contains("=== KeiSei Brain Summary ==="));
|
|
assert!(out.contains("Total DNAs: 3"), "{out}");
|
|
assert!(out.contains("Unique scopes: 2"), "{out}");
|
|
assert!(out.contains("Unique bodies: 3"), "{out}");
|
|
assert!(out.contains("Clusters (scope ≥2):"), "{out}");
|
|
assert!(out.contains("Clusters (body ≥2):"), "{out}");
|
|
assert!(out.contains("Avg cluster size:"), "{out}");
|
|
}
|
|
|
|
#[test]
|
|
fn render_summary_empty_ledger_shows_zeros() {
|
|
let (_d, conn) = seed_db();
|
|
let out = render_summary(&conn).unwrap();
|
|
assert!(out.contains("Total DNAs: 0"), "{out}");
|
|
assert!(out.contains("Unique scopes: 0"), "{out}");
|
|
assert!(out.contains("Unique bodies: 0"), "{out}");
|
|
assert!(out.contains("Clusters (scope ≥2): 0"), "{out}");
|
|
assert!(out.contains("Clusters (body ≥2): 0"), "{out}");
|
|
assert!(out.contains("Avg cluster size: 0.0"), "{out}");
|
|
}
|
|
|
|
#[test]
|
|
fn render_summary_ignores_malformed_dna() {
|
|
let (_d, conn) = seed_db();
|
|
// 3 valid rows + 1 malformed (missing body separator) + 1 NULL.
|
|
let d1 = dna("edit", "NG-FW", "AAAAAAAA", "B1B1B1B1", "11111111");
|
|
let d2 = dna("edit", "NG-FW", "AAAAAAAA", "B2B2B2B2", "22222222");
|
|
let d3 = dna("edit", "NG-FW", "CCCCCCCC", "B3B3B3B3", "33333333");
|
|
insert(&conn, "a1", "running", 1, Some(&d1));
|
|
insert(&conn, "a2", "done", 2, Some(&d2));
|
|
insert(&conn, "a3", "failed", 3, Some(&d3));
|
|
insert(&conn, "bad", "done", 4, Some("not-a-dna"));
|
|
insert(&conn, "nil", "done", 5, None);
|
|
|
|
let out = render_summary(&conn).unwrap();
|
|
// load_rows drops malformed + NULL silently → only 3 valid counted.
|
|
assert!(out.contains("Total DNAs: 3"), "{out}");
|
|
assert!(out.contains("Unique scopes: 2"), "{out}");
|
|
assert!(out.contains("Unique bodies: 3"), "{out}");
|
|
}
|