scai/dist/db/schema.js

> **A local-first AI CLI for understanding, querying, and iterating on large codebases.** > **100% local • No token costs • No cloud • No prompt injection • Private by design**

import { getDbForRepo } from "./client.js";
export function initSchema() {
    const db = getDbForRepo();
    // --- Global state ---
    db.exec(`
    CREATE TABLE IF NOT EXISTS global_state (
      id INTEGER PRIMARY KEY CHECK (id = 1),

      -- user / system preferences
      default_execution_mode TEXT,     -- explain | analyze | transform
      preferred_language TEXT,         -- "en", "da", etc.

      -- learned conventions (lightweight)
      conventions_json TEXT,           -- JSON: naming, formatting, habits
      memory_notes TEXT,               -- freeform distilled memory

      -- bookkeeping
      created_at TEXT NOT NULL,
      updated_at TEXT NOT NULL
    );
  `);
    // --- Projects ---
    db.exec(`
    CREATE TABLE IF NOT EXISTS projects (
      id INTEGER PRIMARY KEY AUTOINCREMENT,

      -- binding (technical, not conceptual)
      repo_path TEXT NOT NULL,          -- absolute path to repo root

      -- human-facing
      name TEXT NOT NULL,               -- project name
      summary TEXT,                     -- evolving description

      -- distilled understanding
      goals_json TEXT,
      constraints_json TEXT,
      assumptions_json TEXT,

      -- bookkeeping
      created_at TEXT NOT NULL,
      updated_at TEXT NOT NULL
    );

    CREATE INDEX IF NOT EXISTS idx_projects_repo
      ON projects(repo_path);
  `);
    // --- Tasks (controlled input only) ---
    db.exec(`
  CREATE TABLE IF NOT EXISTS tasks (
    id INTEGER PRIMARY KEY AUTOINCREMENT,

    project_id INTEGER NOT NULL,

    -- lifecycle
    -- active | completed | paused | needs-feedback
    status TEXT NOT NULL DEFAULT 'active',

    -- user-facing / context
    initial_query TEXT NOT NULL,
    summary TEXT,

    -- resolved intent
    agreed_intent TEXT,
    intent_category TEXT,
    normalized_query TEXT,
    intent_confidence REAL,

    -- focus / file selection
    relevant_files_json TEXT,
    missing_files_json TEXT,

    -- routing decision
    routing_decision_json TEXT,

    -- 🔴 task-level feedback / questions (append-only)
    task_questions_json TEXT,

    -- bookkeeping
    created_at TEXT NOT NULL,
    updated_at TEXT NOT NULL,

    FOREIGN KEY (project_id)
      REFERENCES projects(id)
      ON DELETE CASCADE
  );

  CREATE INDEX IF NOT EXISTS idx_tasks_project
    ON tasks(project_id);

  CREATE INDEX IF NOT EXISTS idx_tasks_status
    ON tasks(status);
`);
    // --- Task steps ---
    db.exec(`
  CREATE TABLE IF NOT EXISTS task_steps (
    id INTEGER PRIMARY KEY AUTOINCREMENT,

    task_id INTEGER NOT NULL,

    file_path TEXT NOT NULL,
    action TEXT,

    -- pending | completed | skipped | needs-feedback
    status TEXT NOT NULL DEFAULT 'pending',

    start_time INTEGER,
    end_time INTEGER,

    result_json TEXT,
    notes TEXT,

    step_index INTEGER,

    -- 🔴 step-level feedback / questions (append-only)
    step_questions_json TEXT,

    created_at TEXT NOT NULL,
    updated_at TEXT NOT NULL,

    FOREIGN KEY (task_id)
      REFERENCES tasks(id)
      ON DELETE CASCADE
  );

  CREATE INDEX IF NOT EXISTS idx_task_steps_task
    ON task_steps(task_id);

  CREATE INDEX IF NOT EXISTS idx_task_steps_status
    ON task_steps(status);

  CREATE UNIQUE INDEX IF NOT EXISTS idx_task_steps_task_file
    ON task_steps(task_id, file_path);
`);
    // --- Core tables ---
    db.exec(`
    CREATE TABLE IF NOT EXISTS files (
      id INTEGER PRIMARY KEY AUTOINCREMENT,
      path TEXT UNIQUE,             -- full path
      filename TEXT,
      summary TEXT,                 -- optional, can be generated later
      content_text TEXT,            -- actual file content
      type TEXT,                    -- file type
      last_modified TEXT,           -- file system last modified
      indexed_at TEXT,              -- timestamp when indexed
      functions_extracted_at TEXT,  -- timestamp when functions/classes extracted
      processing_status TEXT NOT NULL DEFAULT 'unprocessed'  -- tracks pipeline stage
    );

    CREATE VIRTUAL TABLE IF NOT EXISTS files_fts
    USING fts5(
      filename, 
      summary, 
      path, 
      content_text, 
      content='files', 
      content_rowid='id'
    );
  `);
    // --- Folder capsules ---
    db.exec(`
  CREATE TABLE IF NOT EXISTS folder_capsules(
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    path TEXT UNIQUE NOT NULL,
    depth INTEGER NOT NULL,
    capsule_json TEXT NOT NULL,
    confidence REAL,
    last_generated TEXT,
    source_file_count INTEGER
  );

  CREATE INDEX IF NOT EXISTS idx_folder_capsules_path
    ON folder_capsules(path);

  CREATE INDEX IF NOT EXISTS idx_folder_capsules_depth
    ON folder_capsules(depth);
  `);
    // --- Functions table ---
    db.exec(`
    CREATE TABLE IF NOT EXISTS functions(
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    file_id INTEGER REFERENCES files(id),
    name TEXT,
    unique_id TEXT UNIQUE, --e.g. "buildContextualPrompt@cli/src/utils/buildContextualPrompt.ts"
      start_line INTEGER,
    end_line INTEGER,
    content TEXT,
    lang TEXT
  );

    CREATE INDEX IF NOT EXISTS idx_functions_file_id ON functions(file_id);
    CREATE INDEX IF NOT EXISTS idx_functions_unique_id ON functions(unique_id);
  `);
    // --- Graph-specific additions ---
    db.exec(`
    CREATE TABLE IF NOT EXISTS graph_classes(
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    file_id INTEGER REFERENCES files(id),
    name TEXT,
    unique_id TEXT UNIQUE,
    start_line INTEGER,
    end_line INTEGER,
    content TEXT,
    lang TEXT
  );

    CREATE INDEX IF NOT EXISTS idx_graph_classes_file_id ON graph_classes(file_id);
    CREATE INDEX IF NOT EXISTS idx_graph_classes_unique_id ON graph_classes(unique_id);
  `);
    db.exec(`
    CREATE TABLE IF NOT EXISTS graph_edges(
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    source_type TEXT NOT NULL,
    source_unique_id TEXT NOT NULL,
    target_type TEXT NOT NULL,
    target_unique_id TEXT NOT NULL,
    relation TEXT NOT NULL
  );

    CREATE INDEX IF NOT EXISTS idx_graph_edges_source ON graph_edges(source_type, source_unique_id);
    CREATE INDEX IF NOT EXISTS idx_graph_edges_target ON graph_edges(target_type, target_unique_id);
    CREATE INDEX IF NOT EXISTS idx_graph_edges_relation ON graph_edges(relation);
  `);
    db.exec(`
    CREATE TABLE IF NOT EXISTS graph_tags_master(
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    name TEXT UNIQUE NOT NULL
  );

    CREATE TABLE IF NOT EXISTS graph_entity_tags(
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    entity_type TEXT NOT NULL,
    entity_unique_id TEXT NOT NULL,
    tag_id INTEGER NOT NULL REFERENCES graph_tags_master(id),
    UNIQUE(entity_type, entity_unique_id, tag_id)
  );

    CREATE INDEX IF NOT EXISTS idx_graph_entity_tags_entity ON graph_entity_tags(entity_type, entity_unique_id);
    CREATE INDEX IF NOT EXISTS idx_graph_entity_tags_tag ON graph_entity_tags(tag_id);
  `);
    db.exec(`
    CREATE TABLE IF NOT EXISTS summaries(
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    path TEXT UNIQUE,
    type TEXT NOT NULL CHECK(type IN('folder', 'project')),
    summary TEXT,
    last_generated TEXT,
    child_latest_modified TEXT
  );

    CREATE INDEX IF NOT EXISTS idx_summaries_type ON summaries(type);
    CREATE INDEX IF NOT EXISTS idx_summaries_path ON summaries(path);
  `);
    console.log("✅ Graph schema initialized (files, functions, classes, edges, tags, summaries, FTS content_text)");
}