@nostale/cli 0.1.0

package/services/migrations/001_init_rag_v1.sql

@@ -0,0 +1,128 @@
+PRAGMA foreign_keys = ON;
+
+CREATE TABLE documents (
+    id TEXT PRIMARY KEY,
+    project TEXT NOT NULL,
+    repo_group TEXT NOT NULL,
+    doc_type TEXT NOT NULL CHECK (doc_type IN (
+        'product_spec', 'technical_design', 'code_index',
+        'code_module', 'schema', 'graph', 'meta'
+    )),
+    source_path TEXT NOT NULL,
+    title TEXT NOT NULL,
+    module TEXT,
+    feature TEXT,
+    language TEXT,
+    content_hash TEXT NOT NULL,
+    version_tag TEXT,
+    raw_metadata TEXT NOT NULL DEFAULT '{}',
+    ingested_at TEXT NOT NULL DEFAULT (datetime('now')),
+    updated_at TEXT NOT NULL DEFAULT (datetime('now')),
+    UNIQUE (project, source_path, content_hash)
+);
+
+CREATE INDEX documents_project_idx ON documents (project);
+CREATE INDEX documents_doc_type_idx ON documents (doc_type);
+CREATE INDEX documents_module_idx ON documents (module);
+CREATE INDEX documents_feature_idx ON documents (feature);
+
+CREATE TABLE chunks (
+    id TEXT PRIMARY KEY,
+    document_id TEXT NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    chunk_index INTEGER NOT NULL CHECK (chunk_index >= 0),
+    heading_path TEXT NOT NULL DEFAULT '[]',
+    section_title TEXT,
+    content TEXT NOT NULL,
+    content_hash TEXT NOT NULL,
+    token_count INTEGER NOT NULL DEFAULT 0,
+    char_count INTEGER NOT NULL DEFAULT 0,
+    start_line INTEGER,
+    end_line INTEGER,
+    priority INTEGER NOT NULL DEFAULT 50 CHECK (priority BETWEEN 0 AND 100),
+    metadata TEXT NOT NULL DEFAULT '{}',
+    created_at TEXT NOT NULL DEFAULT (datetime('now')),
+    UNIQUE (document_id, chunk_index)
+);
+
+CREATE INDEX chunks_document_id_idx ON chunks (document_id);
+CREATE INDEX chunks_priority_idx ON chunks (priority DESC);
+
+CREATE VIRTUAL TABLE chunks_fts USING fts5(
+    content, section_title, content='chunks', content_rowid='rowid'
+);
+CREATE TRIGGER chunks_fts_ai AFTER INSERT ON chunks BEGIN
+    INSERT INTO chunks_fts(rowid, content, section_title) VALUES (new.rowid, new.content, new.section_title);
+END;
+CREATE TRIGGER chunks_fts_ad AFTER DELETE ON chunks BEGIN
+    INSERT INTO chunks_fts(chunks_fts, rowid, content, section_title) VALUES ('delete', old.rowid, old.content, old.section_title);
+END;
+CREATE TRIGGER chunks_fts_au AFTER UPDATE ON chunks BEGIN
+    INSERT INTO chunks_fts(chunks_fts, rowid, content, section_title) VALUES ('delete', old.rowid, old.content, old.section_title);
+    INSERT INTO chunks_fts(rowid, content, section_title) VALUES (new.rowid, new.content, new.section_title);
+END;
+
+CREATE TABLE embeddings (
+    chunk_id TEXT PRIMARY KEY REFERENCES chunks(id) ON DELETE CASCADE,
+    embedding FLOAT32(1536) NOT NULL,
+    model_name TEXT NOT NULL,
+    dim INTEGER NOT NULL CHECK (dim > 0),
+    embedded_at TEXT NOT NULL DEFAULT (datetime('now'))
+);
+
+CREATE INDEX embeddings_vec_idx ON embeddings (libsql_vector_idx(embedding));
+CREATE INDEX embeddings_model_name_idx ON embeddings (model_name);
+
+CREATE TABLE chunk_entities (
+    id TEXT PRIMARY KEY,
+    chunk_id TEXT NOT NULL REFERENCES chunks(id) ON DELETE CASCADE,
+    entity_type TEXT NOT NULL CHECK (entity_type IN ('symbol', 'route', 'table', 'column')),
+    entity_name TEXT NOT NULL,
+    entity_value TEXT,
+    confidence REAL,
+    created_at TEXT NOT NULL DEFAULT (datetime('now'))
+);
+
+CREATE INDEX chunk_entities_chunk_id_idx ON chunk_entities (chunk_id);
+
+CREATE TABLE graph_nodes (
+    id TEXT PRIMARY KEY,
+    project TEXT NOT NULL,
+    node_type TEXT NOT NULL,
+    node_name TEXT NOT NULL,
+    source_path TEXT,
+    metadata TEXT NOT NULL DEFAULT '{}',
+    UNIQUE (project, node_type, node_name)
+);
+
+CREATE INDEX graph_nodes_source_path_idx ON graph_nodes (project, source_path);
+
+CREATE TABLE graph_edges (
+    id TEXT PRIMARY KEY,
+    project TEXT NOT NULL,
+    from_node_id TEXT NOT NULL REFERENCES graph_nodes(id),
+    to_node_id TEXT NOT NULL REFERENCES graph_nodes(id),
+    edge_type TEXT NOT NULL,
+    weight REAL,
+    source_path TEXT,
+    metadata TEXT NOT NULL DEFAULT '{}',
+    UNIQUE (project, source_path, edge_type, from_node_id, to_node_id)
+);
+
+-- Index berikut menopang query blast-radius incremental sync (lihat 10-incremental-sync.md):
+-- "file mana saja yang mereferensikan simbol dari file yang berubah?"
+CREATE INDEX graph_edges_to_node_id_idx ON graph_edges (to_node_id);
+CREATE INDEX graph_edges_source_path_idx ON graph_edges (project, source_path);
+
+CREATE TABLE retrieval_logs (
+    id TEXT PRIMARY KEY,
+    query_text TEXT NOT NULL,
+    project_filter TEXT,
+    module_filter TEXT,
+    route TEXT NOT NULL,
+    topk_returned INTEGER NOT NULL,
+    candidate_count INTEGER NOT NULL,
+    retrieved_chunk_ids TEXT NOT NULL DEFAULT '[]',
+    answer_confidence REAL,
+    latency_ms INTEGER,
+    created_at TEXT NOT NULL DEFAULT (datetime('now'))
+);

package/services/migrations/002_doc_lifecycle.sql

@@ -0,0 +1,96 @@
+CREATE TABLE source_files (
+    project TEXT NOT NULL,
+    file_path TEXT NOT NULL,
+    content_hash TEXT NOT NULL,
+    updated_at TEXT NOT NULL DEFAULT (datetime('now')),
+    PRIMARY KEY (project, file_path)
+);
+
+CREATE TABLE doc_artifacts (
+    id TEXT PRIMARY KEY,
+    project TEXT NOT NULL,
+    doc_kind TEXT NOT NULL CHECK (doc_kind IN ('prd', 'tdd')),
+    module TEXT NOT NULL,
+    feature TEXT NOT NULL,
+    file_path TEXT NOT NULL,
+    status TEXT NOT NULL DEFAULT 'draft'
+        CHECK (status IN ('draft', 'needs_review', 'approved', 'stale', 'regenerating')),
+    current_version INTEGER NOT NULL DEFAULT 1,
+    source_content_hash TEXT NOT NULL,
+    open_questions_count INTEGER NOT NULL DEFAULT 0,
+    created_at TEXT NOT NULL DEFAULT (datetime('now')),
+    updated_at TEXT NOT NULL DEFAULT (datetime('now')),
+    UNIQUE (project, doc_kind, module, feature)
+);
+
+CREATE INDEX doc_artifacts_project_idx ON doc_artifacts (project);
+CREATE INDEX doc_artifacts_status_idx ON doc_artifacts (status);
+
+CREATE TABLE agent_runs (
+    id TEXT PRIMARY KEY,
+    agent_key TEXT NOT NULL CHECK (agent_key IN (
+        'feature-discovery', 'tdd-writer', 'prd-writer',
+        'documentation-audit', 'gap-summary', 'traceability-matrix'
+    )),
+    project TEXT NOT NULL,
+    target_module TEXT,
+    target_feature TEXT,
+    trigger_source TEXT NOT NULL CHECK (trigger_source IN ('webhook', 'ci', 'manual', 'staleness_scan', 'schedule')),
+    triggered_by_user_id TEXT,
+    status TEXT NOT NULL CHECK (status IN ('running', 'succeeded', 'failed')),
+    llm_provider TEXT NOT NULL,
+    llm_model TEXT NOT NULL,
+    input_token_count INTEGER,
+    output_token_count INTEGER,
+    error_message TEXT,
+    started_at TEXT NOT NULL DEFAULT (datetime('now')),
+    finished_at TEXT
+);
+
+CREATE INDEX agent_runs_project_idx ON agent_runs (project);
+CREATE INDEX agent_runs_started_at_idx ON agent_runs (started_at DESC);
+
+CREATE TABLE doc_versions (
+    id TEXT PRIMARY KEY,
+    doc_artifact_id TEXT NOT NULL REFERENCES doc_artifacts(id) ON DELETE CASCADE,
+    version_number INTEGER NOT NULL,
+    content TEXT NOT NULL,
+    content_hash TEXT NOT NULL,
+    generated_by TEXT NOT NULL,
+    agent_run_id TEXT REFERENCES agent_runs(id),
+    reviewer_note TEXT,
+    reviewed_by_user_id TEXT,
+    reviewed_at TEXT,
+    created_at TEXT NOT NULL DEFAULT (datetime('now')),
+    UNIQUE (doc_artifact_id, version_number)
+);
+
+CREATE INDEX doc_versions_created_at_idx ON doc_versions (created_at);
+
+CREATE TABLE traceability_links (
+    id TEXT PRIMARY KEY,
+    project TEXT NOT NULL,
+    feature TEXT NOT NULL,
+    tdd_artifact_id TEXT REFERENCES doc_artifacts(id),
+    prd_artifact_id TEXT REFERENCES doc_artifacts(id),
+    code_source_paths TEXT NOT NULL DEFAULT '[]',
+    updated_at TEXT NOT NULL DEFAULT (datetime('now'))
+);
+
+CREATE TABLE repo_sync_state (
+    project TEXT PRIMARY KEY,
+    source_type TEXT NOT NULL CHECK (source_type IN ('github_app', 'deploy_key', 'local_path')),
+    git_url TEXT,
+    github_app_installation_id TEXT,
+    deploy_key_secret_ref TEXT,
+    local_path TEXT,
+    default_branch TEXT NOT NULL DEFAULT 'main',
+    latest_known_commit TEXT,
+    last_synced_commit TEXT,
+    last_synced_at TEXT,
+    webhook_secret TEXT,
+    CHECK (
+        (source_type = 'local_path' AND local_path IS NOT NULL)
+        OR (source_type IN ('github_app', 'deploy_key') AND git_url IS NOT NULL)
+    )
+);

package/services/migrations/004_mcp_tokens.sql

@@ -0,0 +1,11 @@
+CREATE TABLE mcp_access_tokens (
+    id TEXT PRIMARY KEY,
+    user_id TEXT NOT NULL,
+    token_hash TEXT NOT NULL UNIQUE,
+    label TEXT,
+    last_used_at TEXT,
+    created_at TEXT NOT NULL DEFAULT (datetime('now')),
+    revoked_at TEXT
+);
+
+CREATE INDEX mcp_access_tokens_user_id_idx ON mcp_access_tokens (user_id);

package/services/migrations/005_settings.sql

@@ -0,0 +1,28 @@
+CREATE TABLE llm_provider_configs (
+    id TEXT PRIMARY KEY,
+    provider TEXT NOT NULL CHECK (provider IN ('anthropic', 'openai_compatible')),
+    label TEXT NOT NULL,
+    api_key_encrypted TEXT NOT NULL,
+    base_url TEXT,
+    default_model TEXT NOT NULL,
+    is_active INTEGER NOT NULL DEFAULT 1 CHECK (is_active IN (0, 1)),
+    created_by_user_id TEXT,
+    created_at TEXT NOT NULL DEFAULT (datetime('now')),
+    updated_at TEXT NOT NULL DEFAULT (datetime('now'))
+);
+
+CREATE TABLE document_templates (
+    id TEXT PRIMARY KEY,
+    project TEXT,
+    doc_kind TEXT NOT NULL CHECK (doc_kind IN ('prd', 'tdd')),
+    name TEXT NOT NULL,
+    sections TEXT NOT NULL,
+    system_prompt_override TEXT,
+    is_active INTEGER NOT NULL DEFAULT 1 CHECK (is_active IN (0, 1)),
+    created_by_user_id TEXT,
+    created_at TEXT NOT NULL DEFAULT (datetime('now')),
+    updated_at TEXT NOT NULL DEFAULT (datetime('now')),
+    UNIQUE (project, doc_kind, name)
+);
+
+CREATE INDEX document_templates_lookup_idx ON document_templates (project, doc_kind, is_active);

package/services/migrations/006_document_sources.sql

@@ -0,0 +1,14 @@
+CREATE TABLE document_sources (
+    id TEXT PRIMARY KEY,
+    project TEXT NOT NULL,
+    source_type TEXT NOT NULL CHECK (source_type IN ('document_folder', 'document_upload')),
+    folder_path TEXT,
+    created_at TEXT NOT NULL DEFAULT (datetime('now')),
+    last_synced_at TEXT,
+    CHECK (
+        (source_type = 'document_folder' AND folder_path IS NOT NULL)
+        OR (source_type = 'document_upload' AND folder_path IS NULL)
+    )
+);
+
+CREATE INDEX document_sources_project_idx ON document_sources (project);

package/services/migrations/007_generation_settings.sql

@@ -0,0 +1,9 @@
+-- M17: per-project toggle for whether sync runs the feature-discovery/tdd-writer/prd-writer
+-- agents after ingestion. Deliberately a new table, not a column on repo_sync_state: a
+-- document-only project (M15/M16's document_sources) never gets a repo_sync_state row, so a
+-- cross-cutting setting living there would be silently inapplicable to exactly the milestone's
+-- own stated primary use case. Same reasoning already used for document_sources in M15.
+CREATE TABLE project_settings (
+    project TEXT PRIMARY KEY,
+    auto_generate_docs INTEGER NOT NULL DEFAULT 1 CHECK (auto_generate_docs IN (0, 1))
+);

package/services/migrations/008_llm_provider_configs_unique.sql

@@ -0,0 +1,5 @@
+-- upsertLlmProviderConfig was a plain INSERT with no natural-key lookup, so every
+-- "edit"/"deactivate" call created a new row instead of updating the existing one
+-- (found during M18's CLI parity work). This unique index lets the fixed upsert use
+-- ON CONFLICT safely, including under concurrent inserts racing the same (provider, label).
+CREATE UNIQUE INDEX llm_provider_configs_provider_label_idx ON llm_provider_configs (provider, label);