sensemaking 0.5.0 → 0.7.0

package/dist/esm/features/embed.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/embed.ts"],"sourcesContent":["import { existsSync, mkdirSync, readFileSync, renameSync, writeFileSync } from 'node:fs';\nimport { homedir } from 'node:os';\nimport { join } from 'node:path';\nimport type { DatabaseSync } from 'node:sqlite';\nimport type { Config, ResolvedConfig } from '../config.ts';\nimport { embedConfig } from '../config.ts';\nimport { SenseError } from '../errors.ts';\nimport { parseFile } from '../scan.ts';\nimport type { Feature } from './types.ts';\n\n// embeddings(path, chunk, start_line, end_line, scale, vector): heading-based chunks,\n// int8 vectors with a per-vector dequantization scale, vector NULL = not yet embedded.\n// Reconcile stores dirty rows inside its transaction; embedding tops up on the next\n// semantic query (embedPending), so staleness costs recall, never correctness.\n\n// Storage lever fixed by the bake-off (BENCHMARKING.md): int8 at 256 dims is\n// quality-free vs f32-512 when fused. Queries stay f32 at the same dims.\nconst STORE_DIMS = 256;\nconst BATCH = 64;\n\nexport interface EmbedProvider {\n  id: string; // model identity; participates in the cache key, change -> re-embed\n  dims: number;\n  embed(texts: string[]): Promise<Float32Array[]>;\n}\n\ninterface Chunk {\n  startLine: number;\n  endLine: number;\n  text: string;\n}\n\n// Deterministic chunker used at reconcile (line ranges stored) and at embed time (text\n// re-derived from the file -- chunk text is never stored). Heading-delimited with the\n// preamble kept; whole file when no headings. The title/summary prefix mirrors the\n// bm25 column weighting.\nfunction chunksOf(raw: string, search?: { title: string; summary: string }): Chunk[] {\n  const lines = raw.split('\\n');\n  const starts: number[] = [];\n  let inFence = false;\n  for (let i = 0; i < lines.length; i++) {\n    if (/^(```|~~~)/.test(lines[i])) {\n      inFence = !inFence;\n      continue;\n    }\n    if (!inFence && /^#{1,6} +/.test(lines[i])) starts.push(i + 1);\n  }\n  const bounds = starts.length === 0 ? [1] : starts[0] > 1 ? [1, ...starts] : starts;\n  const prefix = [search?.title, search?.summary].filter(Boolean).join('\\n');\n  const chunks: Chunk[] = [];\n  bounds.forEach((start, i) => {\n    const end = i + 1 < bounds.length ? bounds[i + 1] - 1 : lines.length;\n    const text = lines\n      .slice(start - 1, end)\n      .join('\\n')\n      .trim();\n    if (text.length > 0) chunks.push({ startLine: start, endLine: end, text: prefix ? `${prefix}\\n${text}` : text });\n  });\n  return chunks;\n}\n\nexport const embed: Feature = {\n  name: 'embed',\n  schema(db) {\n    db.exec(`CREATE TABLE IF NOT EXISTS embeddings (\"path\" TEXT, chunk INTEGER, start_line INTEGER, end_line INTEGER, scale REAL, vector BLOB, PRIMARY KEY (\"path\", chunk))`);\n  },\n  extract(raw, _body, search) {\n    return chunksOf(raw, search);\n  },\n  remove(db, path) {\n    db.prepare('DELETE FROM embeddings WHERE \"path\" = ?').run(path);\n  },\n  store(db, path, extracted) {\n    const insert = db.prepare('INSERT INTO embeddings (\"path\", chunk, start_line, end_line, scale, vector) VALUES (?, ?, ?, ?, NULL, NULL)');\n    (extracted as Chunk[]).forEach((c, idx) => insert.run(path, idx, c.startLine, c.endLine));\n  },\n};\n\n// --- static type: Model2Vec safetensors + pure-JS tokenizer, model cached in ~/.cache ---\n// Encode convention from model2vec/model.py: no special tokens, drop unk ids, mean-pool,\n// L2-normalize.\n\nfunction fetchToFile(url: string, dest: string): Promise<void> {\n  return fetch(url).then(async (res) => {\n    if (!res.ok) throw new SenseError('EMBED_MODEL', `model download failed: ${url} -> HTTP ${res.status}`);\n    writeFileSync(`${dest}.part`, Buffer.from(await res.arrayBuffer()));\n    renameSync(`${dest}.part`, dest);\n  });\n}\n\nasync function staticProvider(model: string): Promise<EmbedProvider> {\n  let dir = model;\n  if (!existsSync(join(dir, 'model.safetensors'))) {\n    dir = join(homedir(), '.cache', 'sensemaking', 'models', model.replace(/\\//g, '--'));\n    mkdirSync(dir, { recursive: true });\n    for (const file of ['model.safetensors', 'tokenizer.json']) {\n      if (!existsSync(join(dir, file))) {\n        console.error(`fetching ${model}/${file} into ${dir} (once; delete to refetch)`);\n        await fetchToFile(`https://huggingface.co/${model}/resolve/main/${file}`, join(dir, file));\n      }\n    }\n  }\n\n  const raw = readFileSync(join(dir, 'model.safetensors'));\n  const headerLen = Number(raw.readBigUInt64LE(0));\n  const header = JSON.parse(raw.subarray(8, 8 + headerLen).toString('utf8')) as Record<string, { dtype: string; shape: number[]; data_offsets: number[] }>;\n  const entry = Object.entries(header).find(([k]) => k !== '__metadata__');\n  if (!entry || entry[1].dtype !== 'F32') throw new SenseError('EMBED_MODEL', `${model}: expected an F32 safetensors matrix`);\n  const spec = entry[1];\n  const dims = spec.shape[1];\n  const dataStart = raw.byteOffset + 8 + headerLen + spec.data_offsets[0];\n  const matrix = dataStart % 4 === 0 ? new Float32Array(raw.buffer, dataStart, spec.shape[0] * dims) : new Float32Array(raw.buffer.slice(dataStart, dataStart + spec.shape[0] * dims * 4));\n\n  const tokenizerJson = JSON.parse(readFileSync(join(dir, 'tokenizer.json'), 'utf8'));\n  // Lazy import: the tokenizer loads only on the semantic path, never at CLI startup.\n  const { Tokenizer } = await import('@huggingface/tokenizers');\n  const tok = new Tokenizer(tokenizerJson, {});\n  const unkId = tokenizerJson.model?.vocab?.[tokenizerJson.model?.unk_token] ?? -1;\n\n  function one(text: string): Float32Array {\n    // The tokenizer yields undefined (not the unk id) for tokens outside the vocab; an\n    // undefined id would index the matrix at NaN and poison the whole mean-pool -- and the\n    // int8 conversion then stores the NaN vector as all zeros, silently. Keep integers only.\n    const ids = (tok.encode(text, { add_special_tokens: false }).ids as number[]).filter((id) => Number.isInteger(id) && id !== unkId);\n    const v = new Float32Array(dims);\n    if (ids.length === 0) return v;\n    for (const id of ids) {\n      const off = id * dims;\n      for (let d = 0; d < dims; d++) v[d] += matrix[off + d];\n    }\n    let norm = 0;\n    for (let d = 0; d < dims; d++) {\n      v[d] /= ids.length;\n      norm += v[d] * v[d];\n    }\n    norm = Math.sqrt(norm) + 1e-32;\n    for (let d = 0; d < dims; d++) v[d] /= norm;\n    return v;\n  }\n\n  return { id: `static:${model}`, dims, embed: async (texts) => texts.map(one) };\n}\n\n// --- api type: one POST against any OpenAI-compatible /embeddings endpoint ---\n\nasync function apiProvider(model: string, url: string | undefined, keyEnv: string | undefined): Promise<EmbedProvider> {\n  if (!url) throw new SenseError('EMBED_MODEL', 'features.embed.type \"api\" requires a url');\n  const base = url.replace(/\\/+$/, '');\n  const headers: Record<string, string> = { 'content-type': 'application/json' };\n  const key = keyEnv ? process.env[keyEnv] : undefined;\n  if (key) headers.authorization = `Bearer ${key}`;\n\n  async function post(texts: string[]): Promise<Float32Array[]> {\n    const res = await fetch(`${base}/embeddings`, { method: 'POST', headers, body: JSON.stringify({ model, input: texts }) });\n    if (!res.ok) throw new SenseError('EMBED_MODEL', `${base}/embeddings -> HTTP ${res.status}`);\n    const body = (await res.json()) as { data: Array<{ embedding: number[] }> };\n    return body.data.map((d) => Float32Array.from(d.embedding));\n  }\n\n  const dims = (await post(['dimension probe']))[0].length;\n  return { id: `api:${base}:${model}`, dims, embed: post };\n}\n\nconst providers = new Map<string, Promise<EmbedProvider>>();\n\nfunction getProvider(cfg: Config): Promise<EmbedProvider> {\n  const e = embedConfig(cfg);\n  if (!e) throw new SenseError('EMBED_DISABLED', 'semantic expansion needs features.embed in sense.config.json (e.g. \"features\": { \"embed\": true })');\n  const sig = `${e.type}:${e.model}:${e.url ?? ''}`;\n  let p = providers.get(sig);\n  if (!p) {\n    p = e.type === 'api' ? apiProvider(e.model, e.url, e.key) : staticProvider(e.model);\n    providers.set(sig, p);\n  }\n  return p;\n}\n\n// Slice + re-normalize (Matryoshka); optionally round through int8 storage.\nfunction toStore(full: Float32Array, dims: number, int8: boolean): { v: Float32Array; scale: number } {\n  const v = new Float32Array(dims);\n  let norm = 0;\n  for (let d = 0; d < dims; d++) norm += full[d] * full[d];\n  norm = Math.sqrt(norm) + 1e-32;\n  for (let d = 0; d < dims; d++) v[d] = full[d] / norm;\n  if (!int8) return { v, scale: 1 };\n  let max = 0;\n  for (let d = 0; d < dims; d++) max = Math.max(max, Math.abs(v[d]));\n  return { v, scale: max / 127 || 1 };\n}\n\n// Embed rows whose vector is NULL, re-deriving chunk text from the files through the\n// same parse + chunker that stored the rows.\nexport async function embedPending(db: DatabaseSync, cfg: Config, baseDir: string): Promise<void> {\n  const provider = await getProvider(cfg); // throws EMBED_DISABLED before touching the table\n  const dirty = db.prepare('SELECT \"path\", chunk FROM embeddings WHERE vector IS NULL ORDER BY \"path\", chunk').all() as Array<{ path: string; chunk: number }>;\n  if (dirty.length === 0) return;\n  const storeDims = Math.min(STORE_DIMS, provider.dims);\n\n  const byPath = new Map<string, number[]>();\n  for (const row of dirty) {\n    const list = byPath.get(row.path) ?? [];\n    list.push(row.chunk);\n    byPath.set(row.path, list);\n  }\n\n  const jobs: Array<{ path: string; chunk: number; text: string }> = [];\n  for (const [path, chunkIdxs] of byPath) {\n    let chunks: Chunk[];\n    try {\n      chunks = parseFile({ relPath: path, absPath: join(baseDir, path), mtimeMs: 0, size: 0 }, [embed]).doc.extracted.embed as Chunk[];\n    } catch {\n      continue; // vanished since reconcile; the next reconcile removes its rows\n    }\n    for (const idx of chunkIdxs) if (chunks[idx]) jobs.push({ path, chunk: idx, text: chunks[idx].text });\n  }\n\n  const update = db.prepare('UPDATE embeddings SET scale = ?, vector = ? WHERE \"path\" = ? AND chunk = ?');\n  for (let i = 0; i < jobs.length; i += BATCH) {\n    const batch = jobs.slice(i, i + BATCH);\n    const vectors = await provider.embed(batch.map((j) => j.text));\n    db.exec('BEGIN');\n    try {\n      batch.forEach((job, j) => {\n        const { v, scale } = toStore(vectors[j], storeDims, true);\n        const q = new Int8Array(storeDims);\n        for (let d = 0; d < storeDims; d++) q[d] = Math.round(v[d] / scale);\n        update.run(scale, Buffer.from(q.buffer), job.path, job.chunk);\n      });\n      db.exec('COMMIT');\n    } catch (err) {\n      db.exec('ROLLBACK');\n      throw err;\n    }\n  }\n}\n\n// Top candidates by cosine for a semantic find: best chunk per file, its line range\n// riding along. FTS5 operators in the terms are lexical syntax, not meaning -- stripped\n// before embedding.\n// Returns the cosine similarity alongside each candidate: stored and query vectors are both\n// L2-normalised before quantisation, so the dot product is cosine in [-1, 1]. `find` surfaces\n// it because the fused rank score cannot express match quality -- nearest-neighbour search\n// always returns a nearest neighbour, so without a magnitude an agent cannot tell a real hit\n// from the best of a bad lot.\nexport async function semanticCandidates(db: DatabaseSync, cfg: Config, terms: string, fetch: number): Promise<Array<{ path: string; lines: string; similarity: number }>> {\n  const baseDir = (cfg as Partial<ResolvedConfig>).baseDir;\n  if (!baseDir) throw new SenseError('EMBED_MODEL', 'semantic expansion needs a config with baseDir (use loadConfig/open)');\n  await embedPending(db, cfg, baseDir);\n\n  const provider = await getProvider(cfg);\n  const storeDims = Math.min(STORE_DIMS, provider.dims);\n  const text = (terms.match(/[\\p{L}\\p{N}]+/gu) ?? []).filter((t) => !['AND', 'OR', 'NOT', 'NEAR'].includes(t)).join(' ');\n  const { v: qv } = toStore((await provider.embed([text]))[0], storeDims, false);\n\n  const rows = db.prepare('SELECT \"path\", start_line, end_line, scale, vector FROM embeddings WHERE vector IS NOT NULL').all() as Array<{\n    path: string;\n    start_line: number;\n    end_line: number;\n    scale: number;\n    vector: Uint8Array;\n  }>;\n\n  const best = new Map<string, { score: number; lines: string }>();\n  for (const row of rows) {\n    const q = new Int8Array(row.vector.buffer, row.vector.byteOffset, Math.min(storeDims, row.vector.byteLength));\n    let dot = 0;\n    for (let d = 0; d < q.length; d++) dot += q[d] * qv[d];\n    const score = dot * row.scale;\n    const existing = best.get(row.path);\n    if (!existing || score > existing.score) best.set(row.path, { score, lines: `L${row.start_line}-${row.end_line}` });\n  }\n  return [...best.entries()]\n    .sort((a, b) => b[1].score - a[1].score)\n    .slice(0, fetch)\n    .map(([path, b]) => ({ path, lines: b.lines, similarity: Math.round(b.score * 1000) / 1000 }));\n}\n"],"names":["existsSync","mkdirSync","readFileSync","renameSync","writeFileSync","homedir","join","embedConfig","SenseError","parseFile","STORE_DIMS","BATCH","chunksOf","raw","search","lines","split","starts","inFence","i","length","test","push","bounds","prefix","title","summary","filter","Boolean","chunks","forEach","start","end","text","slice","trim","startLine","endLine","embed","name","schema","db","exec","extract","_body","remove","path","prepare","run","store","extracted","insert","c","idx","fetchToFile","url","dest","fetch","then","res","ok","status","Buffer","from","arrayBuffer","staticProvider","model","tokenizerJson","dir","replace","recursive","file","console","error","headerLen","Number","readBigUInt64LE","header","JSON","parse","subarray","toString","entry","Object","entries","find","k","dtype","spec","dims","shape","dataStart","byteOffset","data_offsets","matrix","Float32Array","buffer","Tokenizer","tok","unkId","vocab","unk_token","one","ids","encode","add_special_tokens","id","isInteger","v","off","d","norm","Math","sqrt","texts","map","apiProvider","keyEnv","base","headers","key","process","env","undefined","authorization","post","method","body","stringify","input","json","data","embedding","providers","Map","getProvider","cfg","e","sig","type","p","get","set","toStore","full","int8","scale","max","abs","embedPending","baseDir","provider","dirty","all","storeDims","min","byPath","row","list","chunk","jobs","chunkIdxs","relPath","absPath","mtimeMs","size","doc","update","batch","vectors","j","job","q","Int8Array","round","err","semanticCandidates","terms","match","t","includes","qv","rows","best","vector","byteLength","dot","score","existing","start_line","end_line","sort","a","b","similarity"],"mappings":"AAAA,SAASA,UAAU,EAAEC,SAAS,EAAEC,YAAY,EAAEC,UAAU,EAAEC,aAAa,QAAQ,UAAU;AACzF,SAASC,OAAO,QAAQ,UAAU;AAClC,SAASC,IAAI,QAAQ,YAAY;AAGjC,SAASC,WAAW,QAAQ,eAAe;AAC3C,SAASC,UAAU,QAAQ,eAAe;AAC1C,SAASC,SAAS,QAAQ,aAAa;AAGvC,sFAAsF;AACtF,uFAAuF;AACvF,oFAAoF;AACpF,+EAA+E;AAE/E,6EAA6E;AAC7E,yEAAyE;AACzE,MAAMC,aAAa;AACnB,MAAMC,QAAQ;AAcd,uFAAuF;AACvF,sFAAsF;AACtF,mFAAmF;AACnF,yBAAyB;AACzB,SAASC,SAASC,GAAW,EAAEC,MAA2C;IACxE,MAAMC,QAAQF,IAAIG,KAAK,CAAC;IACxB,MAAMC,SAAmB,EAAE;IAC3B,IAAIC,UAAU;IACd,IAAK,IAAIC,IAAI,GAAGA,IAAIJ,MAAMK,MAAM,EAAED,IAAK;QACrC,IAAI,aAAaE,IAAI,CAACN,KAAK,CAACI,EAAE,GAAG;YAC/BD,UAAU,CAACA;YACX;QACF;QACA,IAAI,CAACA,WAAW,YAAYG,IAAI,CAACN,KAAK,CAACI,EAAE,GAAGF,OAAOK,IAAI,CAACH,IAAI;IAC9D;IACA,MAAMI,SAASN,OAAOG,MAAM,KAAK,IAAI;QAAC;KAAE,GAAGH,MAAM,CAAC,EAAE,GAAG,IAAI;QAAC;WAAMA;KAAO,GAAGA;IAC5E,MAAMO,SAAS;QAACV,mBAAAA,6BAAAA,OAAQW,KAAK;QAAEX,mBAAAA,6BAAAA,OAAQY,OAAO;KAAC,CAACC,MAAM,CAACC,SAAStB,IAAI,CAAC;IACrE,MAAMuB,SAAkB,EAAE;IAC1BN,OAAOO,OAAO,CAAC,CAACC,OAAOZ;QACrB,MAAMa,MAAMb,IAAI,IAAII,OAAOH,MAAM,GAAGG,MAAM,CAACJ,IAAI,EAAE,GAAG,IAAIJ,MAAMK,MAAM;QACpE,MAAMa,OAAOlB,MACVmB,KAAK,CAACH,QAAQ,GAAGC,KACjB1B,IAAI,CAAC,MACL6B,IAAI;QACP,IAAIF,KAAKb,MAAM,GAAG,GAAGS,OAAOP,IAAI,CAAC;YAAEc,WAAWL;YAAOM,SAASL;YAAKC,MAAMT,SAAS,GAAGA,OAAO,EAAE,EAAES,MAAM,GAAGA;QAAK;IAChH;IACA,OAAOJ;AACT;AAEA,OAAO,MAAMS,QAAiB;IAC5BC,MAAM;IACNC,QAAOC,EAAE;QACPA,GAAGC,IAAI,CAAC,CAAC,8JAA8J,CAAC;IAC1K;IACAC,SAAQ9B,GAAG,EAAE+B,KAAK,EAAE9B,MAAM;QACxB,OAAOF,SAASC,KAAKC;IACvB;IACA+B,QAAOJ,EAAE,EAAEK,IAAI;QACbL,GAAGM,OAAO,CAAC,2CAA2CC,GAAG,CAACF;IAC5D;IACAG,OAAMR,EAAE,EAAEK,IAAI,EAAEI,SAAS;QACvB,MAAMC,SAASV,GAAGM,OAAO,CAAC;QACzBG,UAAsBpB,OAAO,CAAC,CAACsB,GAAGC,MAAQF,OAAOH,GAAG,CAACF,MAAMO,KAAKD,EAAEhB,SAAS,EAAEgB,EAAEf,OAAO;IACzF;AACF,EAAE;AAEF,2FAA2F;AAC3F,yFAAyF;AACzF,gBAAgB;AAEhB,SAASiB,YAAYC,GAAW,EAAEC,IAAY;IAC5C,OAAOC,MAAMF,KAAKG,IAAI,CAAC,OAAOC;QAC5B,IAAI,CAACA,IAAIC,EAAE,EAAE,MAAM,IAAIpD,WAAW,eAAe,CAAC,uBAAuB,EAAE+C,IAAI,SAAS,EAAEI,IAAIE,MAAM,EAAE;QACtGzD,cAAc,GAAGoD,KAAK,KAAK,CAAC,EAAEM,OAAOC,IAAI,CAAC,MAAMJ,IAAIK,WAAW;QAC/D7D,WAAW,GAAGqD,KAAK,KAAK,CAAC,EAAEA;IAC7B;AACF;AAEA,eAAeS,eAAeC,KAAa;;QA2BEC,sBAA7BA,4BAAAA;IA1Bd,IAAIC,MAAMF;IACV,IAAI,CAAClE,WAAWM,KAAK8D,KAAK,uBAAuB;QAC/CA,MAAM9D,KAAKD,WAAW,UAAU,eAAe,UAAU6D,MAAMG,OAAO,CAAC,OAAO;QAC9EpE,UAAUmE,KAAK;YAAEE,WAAW;QAAK;QACjC,KAAK,MAAMC,QAAQ;YAAC;YAAqB;SAAiB,CAAE;YAC1D,IAAI,CAACvE,WAAWM,KAAK8D,KAAKG,QAAQ;gBAChCC,QAAQC,KAAK,CAAC,CAAC,SAAS,EAAEP,MAAM,CAAC,EAAEK,KAAK,MAAM,EAAEH,IAAI,0BAA0B,CAAC;gBAC/E,MAAMd,YAAY,CAAC,uBAAuB,EAAEY,MAAM,cAAc,EAAEK,MAAM,EAAEjE,KAAK8D,KAAKG;YACtF;QACF;IACF;IAEA,MAAM1D,MAAMX,aAAaI,KAAK8D,KAAK;IACnC,MAAMM,YAAYC,OAAO9D,IAAI+D,eAAe,CAAC;IAC7C,MAAMC,SAASC,KAAKC,KAAK,CAAClE,IAAImE,QAAQ,CAAC,GAAG,IAAIN,WAAWO,QAAQ,CAAC;IAClE,MAAMC,QAAQC,OAAOC,OAAO,CAACP,QAAQQ,IAAI,CAAC,CAAC,CAACC,EAAE,GAAKA,MAAM;IACzD,IAAI,CAACJ,SAASA,KAAK,CAAC,EAAE,CAACK,KAAK,KAAK,OAAO,MAAM,IAAI/E,WAAW,eAAe,GAAG0D,MAAM,oCAAoC,CAAC;IAC1H,MAAMsB,OAAON,KAAK,CAAC,EAAE;IACrB,MAAMO,OAAOD,KAAKE,KAAK,CAAC,EAAE;IAC1B,MAAMC,YAAY9E,IAAI+E,UAAU,GAAG,IAAIlB,YAAYc,KAAKK,YAAY,CAAC,EAAE;IACvE,MAAMC,SAASH,YAAY,MAAM,IAAI,IAAII,aAAalF,IAAImF,MAAM,EAAEL,WAAWH,KAAKE,KAAK,CAAC,EAAE,GAAGD,QAAQ,IAAIM,aAAalF,IAAImF,MAAM,CAAC9D,KAAK,CAACyD,WAAWA,YAAYH,KAAKE,KAAK,CAAC,EAAE,GAAGD,OAAO;IAErL,MAAMtB,gBAAgBW,KAAKC,KAAK,CAAC7E,aAAaI,KAAK8D,KAAK,mBAAmB;IAC3E,oFAAoF;IACpF,MAAM,EAAE6B,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC;IACnC,MAAMC,MAAM,IAAID,UAAU9B,eAAe,CAAC;IAC1C,MAAMgC,iBAAQhC,wBAAAA,cAAcD,KAAK,cAAnBC,6CAAAA,6BAAAA,sBAAqBiC,KAAK,cAA1BjC,iDAAAA,0BAA4B,EAACA,uBAAAA,cAAcD,KAAK,cAAnBC,2CAAAA,qBAAqBkC,SAAS,CAAC,uCAAI,CAAC;IAE/E,SAASC,IAAIrE,IAAY;QACvB,mFAAmF;QACnF,uFAAuF;QACvF,yFAAyF;QACzF,MAAMsE,MAAM,AAACL,IAAIM,MAAM,CAACvE,MAAM;YAAEwE,oBAAoB;QAAM,GAAGF,GAAG,CAAc5E,MAAM,CAAC,CAAC+E,KAAO/B,OAAOgC,SAAS,CAACD,OAAOA,OAAOP;QAC5H,MAAMS,IAAI,IAAIb,aAAaN;QAC3B,IAAIc,IAAInF,MAAM,KAAK,GAAG,OAAOwF;QAC7B,KAAK,MAAMF,MAAMH,IAAK;YACpB,MAAMM,MAAMH,KAAKjB;YACjB,IAAK,IAAIqB,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKF,CAAC,CAACE,EAAE,IAAIhB,MAAM,CAACe,MAAMC,EAAE;QACxD;QACA,IAAIC,OAAO;QACX,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAK;YAC7BF,CAAC,CAACE,EAAE,IAAIP,IAAInF,MAAM;YAClB2F,QAAQH,CAAC,CAACE,EAAE,GAAGF,CAAC,CAACE,EAAE;QACrB;QACAC,OAAOC,KAAKC,IAAI,CAACF,QAAQ;QACzB,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKF,CAAC,CAACE,EAAE,IAAIC;QACvC,OAAOH;IACT;IAEA,OAAO;QAAEF,IAAI,CAAC,OAAO,EAAExC,OAAO;QAAEuB;QAAMnD,OAAO,OAAO4E,QAAUA,MAAMC,GAAG,CAACb;IAAK;AAC/E;AAEA,gFAAgF;AAEhF,eAAec,YAAYlD,KAAa,EAAEX,GAAuB,EAAE8D,MAA0B;IAC3F,IAAI,CAAC9D,KAAK,MAAM,IAAI/C,WAAW,eAAe;IAC9C,MAAM8G,OAAO/D,IAAIc,OAAO,CAAC,QAAQ;IACjC,MAAMkD,UAAkC;QAAE,gBAAgB;IAAmB;IAC7E,MAAMC,MAAMH,SAASI,QAAQC,GAAG,CAACL,OAAO,GAAGM;IAC3C,IAAIH,KAAKD,QAAQK,aAAa,GAAG,CAAC,OAAO,EAAEJ,KAAK;IAEhD,eAAeK,KAAKX,KAAe;QACjC,MAAMvD,MAAM,MAAMF,MAAM,GAAG6D,KAAK,WAAW,CAAC,EAAE;YAAEQ,QAAQ;YAAQP;YAASQ,MAAMjD,KAAKkD,SAAS,CAAC;gBAAE9D;gBAAO+D,OAAOf;YAAM;QAAG;QACvH,IAAI,CAACvD,IAAIC,EAAE,EAAE,MAAM,IAAIpD,WAAW,eAAe,GAAG8G,KAAK,oBAAoB,EAAE3D,IAAIE,MAAM,EAAE;QAC3F,MAAMkE,OAAQ,MAAMpE,IAAIuE,IAAI;QAC5B,OAAOH,KAAKI,IAAI,CAAChB,GAAG,CAAC,CAACL,IAAMf,aAAahC,IAAI,CAAC+C,EAAEsB,SAAS;IAC3D;IAEA,MAAM3C,OAAO,AAAC,CAAA,MAAMoC,KAAK;QAAC;KAAkB,CAAA,CAAE,CAAC,EAAE,CAACzG,MAAM;IACxD,OAAO;QAAEsF,IAAI,CAAC,IAAI,EAAEY,KAAK,CAAC,EAAEpD,OAAO;QAAEuB;QAAMnD,OAAOuF;IAAK;AACzD;AAEA,MAAMQ,YAAY,IAAIC;AAEtB,SAASC,YAAYC,GAAW;QAGMC;IAFpC,MAAMA,IAAIlI,YAAYiI;IACtB,IAAI,CAACC,GAAG,MAAM,IAAIjI,WAAW,kBAAkB;IAC/C,MAAMkI,MAAM,GAAGD,EAAEE,IAAI,CAAC,CAAC,EAAEF,EAAEvE,KAAK,CAAC,CAAC,GAAEuE,SAAAA,EAAElF,GAAG,cAALkF,oBAAAA,SAAS,IAAI;IACjD,IAAIG,IAAIP,UAAUQ,GAAG,CAACH;IACtB,IAAI,CAACE,GAAG;QACNA,IAAIH,EAAEE,IAAI,KAAK,QAAQvB,YAAYqB,EAAEvE,KAAK,EAAEuE,EAAElF,GAAG,EAAEkF,EAAEjB,GAAG,IAAIvD,eAAewE,EAAEvE,KAAK;QAClFmE,UAAUS,GAAG,CAACJ,KAAKE;IACrB;IACA,OAAOA;AACT;AAEA,4EAA4E;AAC5E,SAASG,QAAQC,IAAkB,EAAEvD,IAAY,EAAEwD,IAAa;IAC9D,MAAMrC,IAAI,IAAIb,aAAaN;IAC3B,IAAIsB,OAAO;IACX,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKC,QAAQiC,IAAI,CAAClC,EAAE,GAAGkC,IAAI,CAAClC,EAAE;IACxDC,OAAOC,KAAKC,IAAI,CAACF,QAAQ;IACzB,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKF,CAAC,CAACE,EAAE,GAAGkC,IAAI,CAAClC,EAAE,GAAGC;IAChD,IAAI,CAACkC,MAAM,OAAO;QAAErC;QAAGsC,OAAO;IAAE;IAChC,IAAIC,MAAM;IACV,IAAK,IAAIrC,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKqC,MAAMnC,KAAKmC,GAAG,CAACA,KAAKnC,KAAKoC,GAAG,CAACxC,CAAC,CAACE,EAAE;IAChE,OAAO;QAAEF;QAAGsC,OAAOC,MAAM,OAAO;IAAE;AACpC;AAEA,qFAAqF;AACrF,6CAA6C;AAC7C,OAAO,eAAeE,aAAa5G,EAAgB,EAAE+F,GAAW,EAAEc,OAAe;IAC/E,MAAMC,WAAW,MAAMhB,YAAYC,MAAM,kDAAkD;IAC3F,MAAMgB,QAAQ/G,GAAGM,OAAO,CAAC,oFAAoF0G,GAAG;IAChH,IAAID,MAAMpI,MAAM,KAAK,GAAG;IACxB,MAAMsI,YAAY1C,KAAK2C,GAAG,CAACjJ,YAAY6I,SAAS9D,IAAI;IAEpD,MAAMmE,SAAS,IAAItB;IACnB,KAAK,MAAMuB,OAAOL,MAAO;YACVI;QAAb,MAAME,QAAOF,cAAAA,OAAOf,GAAG,CAACgB,IAAI/G,IAAI,eAAnB8G,yBAAAA,cAAwB,EAAE;QACvCE,KAAKxI,IAAI,CAACuI,IAAIE,KAAK;QACnBH,OAAOd,GAAG,CAACe,IAAI/G,IAAI,EAAEgH;IACvB;IAEA,MAAME,OAA6D,EAAE;IACrE,KAAK,MAAM,CAAClH,MAAMmH,UAAU,IAAIL,OAAQ;QACtC,IAAI/H;QACJ,IAAI;YACFA,SAASpB,UAAU;gBAAEyJ,SAASpH;gBAAMqH,SAAS7J,KAAKgJ,SAASxG;gBAAOsH,SAAS;gBAAGC,MAAM;YAAE,GAAG;gBAAC/H;aAAM,EAAEgI,GAAG,CAACpH,SAAS,CAACZ,KAAK;QACvH,EAAE,OAAM;YACN,UAAU,gEAAgE;QAC5E;QACA,KAAK,MAAMe,OAAO4G,UAAW,IAAIpI,MAAM,CAACwB,IAAI,EAAE2G,KAAK1I,IAAI,CAAC;YAAEwB;YAAMiH,OAAO1G;YAAKpB,MAAMJ,MAAM,CAACwB,IAAI,CAACpB,IAAI;QAAC;IACrG;IAEA,MAAMsI,SAAS9H,GAAGM,OAAO,CAAC;IAC1B,IAAK,IAAI5B,IAAI,GAAGA,IAAI6I,KAAK5I,MAAM,EAAED,KAAKR,MAAO;QAC3C,MAAM6J,QAAQR,KAAK9H,KAAK,CAACf,GAAGA,IAAIR;QAChC,MAAM8J,UAAU,MAAMlB,SAASjH,KAAK,CAACkI,MAAMrD,GAAG,CAAC,CAACuD,IAAMA,EAAEzI,IAAI;QAC5DQ,GAAGC,IAAI,CAAC;QACR,IAAI;YACF8H,MAAM1I,OAAO,CAAC,CAAC6I,KAAKD;gBAClB,MAAM,EAAE9D,CAAC,EAAEsC,KAAK,EAAE,GAAGH,QAAQ0B,OAAO,CAACC,EAAE,EAAEhB,WAAW;gBACpD,MAAMkB,IAAI,IAAIC,UAAUnB;gBACxB,IAAK,IAAI5C,IAAI,GAAGA,IAAI4C,WAAW5C,IAAK8D,CAAC,CAAC9D,EAAE,GAAGE,KAAK8D,KAAK,CAAClE,CAAC,CAACE,EAAE,GAAGoC;gBAC7DqB,OAAOvH,GAAG,CAACkG,OAAOpF,OAAOC,IAAI,CAAC6G,EAAE5E,MAAM,GAAG2E,IAAI7H,IAAI,EAAE6H,IAAIZ,KAAK;YAC9D;YACAtH,GAAGC,IAAI,CAAC;QACV,EAAE,OAAOqI,KAAK;YACZtI,GAAGC,IAAI,CAAC;YACR,MAAMqI;QACR;IACF;AACF;AAEA,oFAAoF;AACpF,wFAAwF;AACxF,oBAAoB;AACpB,4FAA4F;AAC5F,8FAA8F;AAC9F,2FAA2F;AAC3F,6FAA6F;AAC7F,8BAA8B;AAC9B,OAAO,eAAeC,mBAAmBvI,EAAgB,EAAE+F,GAAW,EAAEyC,KAAa,EAAExH,MAAa;QAOpFwH;IANd,MAAM3B,UAAU,AAACd,IAAgCc,OAAO;IACxD,IAAI,CAACA,SAAS,MAAM,IAAI9I,WAAW,eAAe;IAClD,MAAM6I,aAAa5G,IAAI+F,KAAKc;IAE5B,MAAMC,WAAW,MAAMhB,YAAYC;IACnC,MAAMkB,YAAY1C,KAAK2C,GAAG,CAACjJ,YAAY6I,SAAS9D,IAAI;IACpD,MAAMxD,OAAO,EAACgJ,eAAAA,MAAMC,KAAK,CAAC,gCAAZD,0BAAAA,eAAkC,EAAE,EAAEtJ,MAAM,CAAC,CAACwJ,IAAM,CAAC;YAAC;YAAO;YAAM;YAAO;SAAO,CAACC,QAAQ,CAACD,IAAI7K,IAAI,CAAC;IAClH,MAAM,EAAEsG,GAAGyE,EAAE,EAAE,GAAGtC,QAAQ,AAAC,CAAA,MAAMQ,SAASjH,KAAK,CAAC;QAACL;KAAK,CAAA,CAAE,CAAC,EAAE,EAAEyH,WAAW;IAExE,MAAM4B,OAAO7I,GAAGM,OAAO,CAAC,+FAA+F0G,GAAG;IAQ1H,MAAM8B,OAAO,IAAIjD;IACjB,KAAK,MAAMuB,OAAOyB,KAAM;QACtB,MAAMV,IAAI,IAAIC,UAAUhB,IAAI2B,MAAM,CAACxF,MAAM,EAAE6D,IAAI2B,MAAM,CAAC5F,UAAU,EAAEoB,KAAK2C,GAAG,CAACD,WAAWG,IAAI2B,MAAM,CAACC,UAAU;QAC3G,IAAIC,MAAM;QACV,IAAK,IAAI5E,IAAI,GAAGA,IAAI8D,EAAExJ,MAAM,EAAE0F,IAAK4E,OAAOd,CAAC,CAAC9D,EAAE,GAAGuE,EAAE,CAACvE,EAAE;QACtD,MAAM6E,QAAQD,MAAM7B,IAAIX,KAAK;QAC7B,MAAM0C,WAAWL,KAAK1C,GAAG,CAACgB,IAAI/G,IAAI;QAClC,IAAI,CAAC8I,YAAYD,QAAQC,SAASD,KAAK,EAAEJ,KAAKzC,GAAG,CAACe,IAAI/G,IAAI,EAAE;YAAE6I;YAAO5K,OAAO,CAAC,CAAC,EAAE8I,IAAIgC,UAAU,CAAC,CAAC,EAAEhC,IAAIiC,QAAQ,EAAE;QAAC;IACnH;IACA,OAAO;WAAIP,KAAKnG,OAAO;KAAG,CACvB2G,IAAI,CAAC,CAACC,GAAGC,IAAMA,CAAC,CAAC,EAAE,CAACN,KAAK,GAAGK,CAAC,CAAC,EAAE,CAACL,KAAK,EACtCzJ,KAAK,CAAC,GAAGuB,QACT0D,GAAG,CAAC,CAAC,CAACrE,MAAMmJ,EAAE,GAAM,CAAA;YAAEnJ;YAAM/B,OAAOkL,EAAElL,KAAK;YAAEmL,YAAYlF,KAAK8D,KAAK,CAACmB,EAAEN,KAAK,GAAG,QAAQ;QAAK,CAAA;AAC/F"}

package/dist/esm/features/index.js

@@ -1,4 +1,5 @@
 import { featureEnabled } from '../config.js';
+import { embed } from './embed.js';
 import { links } from './links.js';
 import { rank } from './rank.js';
 import { sections } from './sections.js';
@@ -6,7 +7,8 @@ import { sections } from './sections.js';
 export const FEATURES = [
     links,
     sections,
-    rank
+    rank,
+    embed
 ];
 export function activeFeatures(cfg) {
     return FEATURES.filter((feature)=>featureEnabled(cfg, feature.name));

package/dist/esm/features/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/index.ts"],"sourcesContent":["import type { Config } from '../config.ts';\nimport { featureEnabled } from '../config.ts';\nimport { links } from './links.ts';\nimport { rank } from './rank.ts';\nimport { sections } from './sections.ts';\nimport type { Feature } from './types.ts';\n\n// Registry order matters: rank reads the links table in afterReconcile.\nexport const FEATURES: Feature[] = [links, sections, rank];\n\nexport function activeFeatures(cfg: Config): Feature[] {\n  return FEATURES.filter((feature) => featureEnabled(cfg, feature.name));\n}\n\nexport { linkEdges } from './links.ts';\nexport type { Section } from './sections.ts';\nexport type { Feature } from './types.ts';\n"],"names":["featureEnabled","links","rank","sections","FEATURES","activeFeatures","cfg","filter","feature","name","linkEdges"],"mappings":"AACA,SAASA,cAAc,QAAQ,eAAe;AAC9C,SAASC,KAAK,QAAQ,aAAa;AACnC,SAASC,IAAI,QAAQ,YAAY;AACjC,SAASC,QAAQ,QAAQ,gBAAgB;AAGzC,wEAAwE;AACxE,OAAO,MAAMC,WAAsB;IAACH;IAAOE;IAAUD;CAAK,CAAC;AAE3D,OAAO,SAASG,eAAeC,GAAW;IACxC,OAAOF,SAASG,MAAM,CAAC,CAACC,UAAYR,eAAeM,KAAKE,QAAQC,IAAI;AACtE;AAEA,SAASC,SAAS,QAAQ,aAAa"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/index.ts"],"sourcesContent":["import type { Config } from '../config.ts';\nimport { featureEnabled } from '../config.ts';\nimport { embed } from './embed.ts';\nimport { links } from './links.ts';\nimport { rank } from './rank.ts';\nimport { sections } from './sections.ts';\nimport type { Feature } from './types.ts';\n\n// Registry order matters: rank reads the links table in afterReconcile.\nexport const FEATURES: Feature[] = [links, sections, rank, embed];\n\nexport function activeFeatures(cfg: Config): Feature[] {\n  return FEATURES.filter((feature) => featureEnabled(cfg, feature.name));\n}\n\nexport { linkEdges } from './links.ts';\nexport type { Section } from './sections.ts';\nexport type { Feature } from './types.ts';\n"],"names":["featureEnabled","embed","links","rank","sections","FEATURES","activeFeatures","cfg","filter","feature","name","linkEdges"],"mappings":"AACA,SAASA,cAAc,QAAQ,eAAe;AAC9C,SAASC,KAAK,QAAQ,aAAa;AACnC,SAASC,KAAK,QAAQ,aAAa;AACnC,SAASC,IAAI,QAAQ,YAAY;AACjC,SAASC,QAAQ,QAAQ,gBAAgB;AAGzC,wEAAwE;AACxE,OAAO,MAAMC,WAAsB;IAACH;IAAOE;IAAUD;IAAMF;CAAM,CAAC;AAElE,OAAO,SAASK,eAAeC,GAAW;IACxC,OAAOF,SAASG,MAAM,CAAC,CAACC,UAAYT,eAAeO,KAAKE,QAAQC,IAAI;AACtE;AAEA,SAASC,SAAS,QAAQ,aAAa"}

package/dist/esm/features/types.d.ts

@@ -4,7 +4,10 @@ import type { FileStat } from '../scan.js';
 export interface Feature {
     name: FeatureName;
     schema(db: DatabaseSync): void;
-    extract?(raw: string, body: string): unknown;
+    extract?(raw: string, body: string, search?: {
+        title: string;
+        summary: string;
+    }): unknown;
     remove?(db: DatabaseSync, path: string): void;
     store?(db: DatabaseSync, path: string, extracted: unknown): void;
     afterReconcile?(db: DatabaseSync, files: FileStat[]): void;

package/dist/esm/features/types.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/types.ts"],"sourcesContent":["import type { DatabaseSync } from 'node:sqlite';\nimport type { FeatureName } from '../config.ts';\nimport type { FileStat } from '../scan.ts';\n\n// A feature owns its schema, per-file extraction/storage, and any whole-tree pass.\n// Adding a feature = one module here + one entry in the registry (index.ts).\nexport interface Feature {\n  name: FeatureName;\n  schema(db: DatabaseSync): void;\n  // Pure, per file: raw is the full file, body the prose after frontmatter.\n  extract?(raw: string, body: string): unknown;\n  remove?(db: DatabaseSync, path: string): void;\n  store?(db: DatabaseSync, path: string, extracted: unknown): void;\n  // After all rows are current, inside the reconcile transaction (resolution, rank).\n  afterReconcile?(db: DatabaseSync, files: FileStat[]): void;\n}\n"],"names":[],"mappings":"AAIA,mFAAmF;AACnF,6EAA6E;AAC7E,WASC"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/types.ts"],"sourcesContent":["import type { DatabaseSync } from 'node:sqlite';\nimport type { FeatureName } from '../config.ts';\nimport type { FileStat } from '../scan.ts';\n\n// A feature owns its schema, per-file extraction/storage, and any whole-tree pass.\n// Adding a feature = one module here + one entry in the registry (index.ts).\nexport interface Feature {\n  name: FeatureName;\n  schema(db: DatabaseSync): void;\n  // Pure, per file: raw is the full file, body the prose after frontmatter,\n  // search the normalized title/summary strings.\n  extract?(raw: string, body: string, search?: { title: string; summary: string }): unknown;\n  remove?(db: DatabaseSync, path: string): void;\n  store?(db: DatabaseSync, path: string, extracted: unknown): void;\n  // After all rows are current, inside the reconcile transaction (resolution, rank).\n  afterReconcile?(db: DatabaseSync, files: FileStat[]): void;\n}\n"],"names":[],"mappings":"AAIA,mFAAmF;AACnF,6EAA6E;AAC7E,WAUC"}

package/dist/esm/index.d.ts

@@ -1,4 +1,4 @@
-export type { Config, FeatureName, ResolvedConfig } from './config.js';
+export type { Config, EmbedConfig, FeatureName, ResolvedConfig } from './config.js';
 export { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.js';
 export type { OpenResult } from './db.js';
 export { open, rebuild } from './db.js';

package/dist/esm/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/index.ts"],"sourcesContent":["// Public library API. Deliberately small: every export is a stability promise;\n// internals (feature registry, graph, scan, meta) stay module-private.\n\nexport type { Config, FeatureName, ResolvedConfig } from './config.ts';\nexport { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.ts';\n\nexport type { OpenResult } from './db.ts';\nexport { open, rebuild } from './db.ts';\n\nexport type { SenseErrorCode } from './errors.ts';\nexport { SenseError } from './errors.ts';\n\nexport type { Row } from './output.ts';\nexport { printRows } from './output.ts';\n\nexport type { FindOptions, Peek, TreeMap } from './verbs.ts';\nexport { find, mapTree, peek } from './verbs.ts';\n\nexport type { WatchEvent, WatchOptions } from './watch.ts';\nexport { runWatch } from './watch.ts';\n"],"names":["CONFIG_FILENAME","initConfig","loadConfig","migrateConfig","STATE_DIR","SUPPORTED_CONFIG_VERSION","open","rebuild","SenseError","printRows","find","mapTree","peek","runWatch"],"mappings":"AAAA,+EAA+E;AAC/E,uEAAuE;AAGvE,SAASA,eAAe,EAAEC,UAAU,EAAEC,UAAU,EAAEC,aAAa,EAAEC,SAAS,EAAEC,wBAAwB,QAAQ,cAAc;AAG1H,SAASC,IAAI,EAAEC,OAAO,QAAQ,UAAU;AAGxC,SAASC,UAAU,QAAQ,cAAc;AAGzC,SAASC,SAAS,QAAQ,cAAc;AAGxC,SAASC,IAAI,EAAEC,OAAO,EAAEC,IAAI,QAAQ,aAAa;AAGjD,SAASC,QAAQ,QAAQ,aAAa"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/index.ts"],"sourcesContent":["// Public library API. Deliberately small: every export is a stability promise;\n// internals (feature registry, graph, scan, meta) stay module-private.\n\nexport type { Config, EmbedConfig, FeatureName, ResolvedConfig } from './config.ts';\nexport { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.ts';\n\nexport type { OpenResult } from './db.ts';\nexport { open, rebuild } from './db.ts';\n\nexport type { SenseErrorCode } from './errors.ts';\nexport { SenseError } from './errors.ts';\n\nexport type { Row } from './output.ts';\nexport { printRows } from './output.ts';\n\nexport type { FindOptions, Peek, TreeMap } from './verbs.ts';\nexport { find, mapTree, peek } from './verbs.ts';\n\nexport type { WatchEvent, WatchOptions } from './watch.ts';\nexport { runWatch } from './watch.ts';\n"],"names":["CONFIG_FILENAME","initConfig","loadConfig","migrateConfig","STATE_DIR","SUPPORTED_CONFIG_VERSION","open","rebuild","SenseError","printRows","find","mapTree","peek","runWatch"],"mappings":"AAAA,+EAA+E;AAC/E,uEAAuE;AAGvE,SAASA,eAAe,EAAEC,UAAU,EAAEC,UAAU,EAAEC,aAAa,EAAEC,SAAS,EAAEC,wBAAwB,QAAQ,cAAc;AAG1H,SAASC,IAAI,EAAEC,OAAO,QAAQ,UAAU;AAGxC,SAASC,UAAU,QAAQ,cAAc;AAGzC,SAASC,SAAS,QAAQ,cAAc;AAGxC,SAASC,IAAI,EAAEC,OAAO,EAAEC,IAAI,QAAQ,aAAa;AAGjD,SAASC,QAAQ,QAAQ,aAAa"}

package/dist/esm/output.d.ts

@@ -7,6 +7,10 @@ export declare function renderMap(result: {
     };
     fields: Row[];
     fieldsTotal: number;
+    features: {
+        on: string[];
+        off: string[];
+    };
     hubs: Row[];
     recent: Row[];
 }): string;
@@ -21,4 +25,5 @@ export declare function renderPeek(result: {
     outboundTotal: number;
     backlinksTotal: number;
     unresolvedTotal: number;
+    off: string[];
 }): string;

package/dist/esm/output.js

@@ -6,19 +6,41 @@ function cell(value) {
     return String(value !== null && value !== void 0 ? value : '').replace(/\s*\n\s*/g, ' ');
 }
 export function printRows(rows, format) {
-    const rendered = render(rows, format);
+    const rendered = renderRows(rows, format);
     console.log(rendered);
     if (rendered.length > OVERSIZE_BYTES) {
         console.warn(`warning: result is ${Math.round(rendered.length / 1000)} KB across ${rows.length} row(s); consider a LIMIT, fewer columns, or snippet() instead of whole values`);
     }
 }
-function render(rows, format) {
+// Table output is for reading; a row wider than the terminal wraps into an unreadable
+// block (a `find` header measured 330 chars once `summary` is present, which drove one
+// user to pipe --format json through a script for every interactive query). Widest
+// columns give up space first, down to a floor, and only as far as the terminal needs.
+// --format json is never truncated: it is the machine-readable surface.
+const MIN_COLUMN = 12;
+const TRUNCATED = '…';
+function fitWidths(widths, limit) {
+    const gaps = (widths.length - 1) * 2;
+    const fitted = [
+        ...widths
+    ];
+    for(;;){
+        const total = fitted.reduce((a, b)=>a + b, 0) + gaps;
+        if (total <= limit) return fitted;
+        const widest = fitted.indexOf(Math.max(...fitted));
+        if (fitted[widest] <= MIN_COLUMN) return fitted;
+        fitted[widest] = Math.max(MIN_COLUMN, fitted[widest] - Math.max(1, total - limit));
+    }
+}
+function renderRows(rows, format, width = process.stdout.columns) {
     if (format === 'json') return JSON.stringify(rows, null, 2);
     if (rows.length === 0) return '(0 rows)';
     const columns = Object.keys(rows[0]);
     const cells = rows.map((row)=>columns.map((col)=>cell(row[col])));
-    const widths = columns.map((col, i)=>Math.max(col.length, ...cells.map((row)=>row[i].length)));
-    const formatRow = (values)=>values.map((value, i)=>value.padEnd(widths[i])).join('  ');
+    const natural = columns.map((col, i)=>Math.max(col.length, ...cells.map((row)=>row[i].length)));
+    const widths = width && width > MIN_COLUMN ? fitWidths(natural, width) : natural;
+    const clip = (value, w)=>value.length <= w ? value.padEnd(w) : `${value.slice(0, w - 1)}${TRUNCATED}`;
+    const formatRow = (values)=>values.map((value, i)=>clip(value, widths[i])).join('  ').trimEnd();
     return [
         formatRow(columns),
         widths.map((w)=>'-'.repeat(w)).join('  '),
@@ -26,14 +48,20 @@ function render(rows, format) {
     ].join('\n');
 }
 // Text renderers for the layer verbs; cli.ts prints what these return.
+function featuresLine(features) {
+    const off = features.off.length > 0 ? ` · off: ${features.off.map((name)=>`${name} (features.${name})`).join(', ')}` : '';
+    return `features: ${features.on.join(', ')}${off}`;
+}
 export function renderMap(result) {
     const parts = [
-        `docs: ${result.docs.count} (${Math.round(result.docs.bytes / 1024)} KB)\n`,
-        render(result.fields, 'table')
+        `docs: ${result.docs.count} (${Math.round(result.docs.bytes / 1024)} KB)`,
+        `${featuresLine(result.features)}\n`,
+        renderRows(result.fields, 'table')
     ];
     if (result.fieldsTotal > result.fields.length) parts.push(`(+${result.fieldsTotal - result.fields.length} more fields)`);
-    if (result.hubs.length > 0) parts.push('\nhubs (by link rank):', render(result.hubs, 'table'));
-    parts.push('\nrecent:', render(result.recent, 'table'));
+    if (result.hubs.length > 0) parts.push('\nhubs (by link rank):', renderRows(result.hubs, 'table'));
+    else if (result.features.off.includes('rank')) parts.push('\nhubs: off (features.rank)');
+    parts.push('\nrecent:', renderRows(result.recent, 'table'));
     return parts.join('\n');
 }
 export function renderPeek(result) {
@@ -44,6 +72,10 @@ export function renderPeek(result) {
     if (result.sections.length > 0) {
         lines.push('', 'sections:');
         for (const s of result.sections)lines.push(`  ${'#'.repeat(s.level)} ${s.heading}  [L${s.start_line}-${s.end_line}, ~${s.tokens}t]`);
+    } else if (result.off.includes('sections')) lines.push('', 'sections: off (features.sections)');
+    if (result.off.includes('links')) {
+        lines.push('', 'links: off (features.links)');
+        return lines.join('\n');
     }
     const linkLine = (label, shown, total)=>{
         if (total > 0) lines.push(`${label} (${total}): ${shown.join(', ')}${total > shown.length ? `, +${total - shown.length} more` : ''}`);

package/dist/esm/output.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/output.ts"],"sourcesContent":["// rows -> table (default) | json\n\nexport type Row = Record<string, unknown>;\n\n// Warned about on stderr (keeps --format json stdout machine-readable), never truncated.\nconst OVERSIZE_BYTES = 50_000;\n\n// Flatten embedded newlines so they can't break table column alignment; JSON output is left alone.\nfunction cell(value: unknown): string {\n  return String(value ?? '').replace(/\\s*\\n\\s*/g, ' ');\n}\n\nexport function printRows(rows: Row[], format: 'table' | 'json'): void {\n  const rendered = render(rows, format);\n  console.log(rendered);\n  if (rendered.length > OVERSIZE_BYTES) {\n    console.warn(`warning: result is ${Math.round(rendered.length / 1000)} KB across ${rows.length} row(s); consider a LIMIT, fewer columns, or snippet() instead of whole values`);\n  }\n}\n\nfunction render(rows: Row[], format: 'table' | 'json'): string {\n  if (format === 'json') return JSON.stringify(rows, null, 2);\n  if (rows.length === 0) return '(0 rows)';\n\n  const columns = Object.keys(rows[0]);\n  const cells = rows.map((row) => columns.map((col) => cell(row[col])));\n  const widths = columns.map((col, i) => Math.max(col.length, ...cells.map((row) => row[i].length)));\n\n  const formatRow = (values: string[]) => values.map((value, i) => value.padEnd(widths[i])).join('  ');\n\n  return [formatRow(columns), widths.map((w) => '-'.repeat(w)).join('  '), ...cells.map(formatRow)].join('\\n');\n}\n\n// Text renderers for the layer verbs; cli.ts prints what these return.\n\nexport function renderMap(result: { docs: { count: number; bytes: number }; fields: Row[]; fieldsTotal: number; hubs: Row[]; recent: Row[] }): string {\n  const parts = [`docs: ${result.docs.count} (${Math.round(result.docs.bytes / 1024)} KB)\\n`, render(result.fields, 'table')];\n  if (result.fieldsTotal > result.fields.length) parts.push(`(+${result.fieldsTotal - result.fields.length} more fields)`);\n  if (result.hubs.length > 0) parts.push('\\nhubs (by link rank):', render(result.hubs, 'table'));\n  parts.push('\\nrecent:', render(result.recent, 'table'));\n  return parts.join('\\n');\n}\n\nexport function renderPeek(result: { path: string; tokens: number; frontmatter: Row; sections: Row[]; outbound: string[]; backlinks: string[]; unresolved: string[]; outboundTotal: number; backlinksTotal: number; unresolvedTotal: number }): string {\n  const lines = [`${result.path}  (~${result.tokens} tokens)`];\n  for (const [key, value] of Object.entries(result.frontmatter)) lines.push(`  ${key}: ${value}`);\n  if (result.sections.length > 0) {\n    lines.push('', 'sections:');\n    for (const s of result.sections) lines.push(`  ${'#'.repeat(s.level as number)} ${s.heading}  [L${s.start_line}-${s.end_line}, ~${s.tokens}t]`);\n  }\n  const linkLine = (label: string, shown: string[], total: number) => {\n    if (total > 0) lines.push(`${label} (${total}): ${shown.join(', ')}${total > shown.length ? `, +${total - shown.length} more` : ''}`);\n  };\n  if (result.outboundTotal + result.unresolvedTotal + result.backlinksTotal > 0) lines.push('');\n  linkLine('links out', result.outbound, result.outboundTotal);\n  linkLine('unresolved', result.unresolved, result.unresolvedTotal);\n  linkLine('backlinks', result.backlinks, result.backlinksTotal);\n  return lines.join('\\n');\n}\n"],"names":["OVERSIZE_BYTES","cell","value","String","replace","printRows","rows","format","rendered","render","console","log","length","warn","Math","round","JSON","stringify","columns","Object","keys","cells","map","row","col","widths","i","max","formatRow","values","padEnd","join","w","repeat","renderMap","result","parts","docs","count","bytes","fields","fieldsTotal","push","hubs","recent","renderPeek","lines","path","tokens","key","entries","frontmatter","sections","s","level","heading","start_line","end_line","linkLine","label","shown","total","outboundTotal","unresolvedTotal","backlinksTotal","outbound","unresolved","backlinks"],"mappings":"AAAA,iCAAiC;AAIjC,yFAAyF;AACzF,MAAMA,iBAAiB;AAEvB,mGAAmG;AACnG,SAASC,KAAKC,KAAc;IAC1B,OAAOC,OAAOD,kBAAAA,mBAAAA,QAAS,IAAIE,OAAO,CAAC,aAAa;AAClD;AAEA,OAAO,SAASC,UAAUC,IAAW,EAAEC,MAAwB;IAC7D,MAAMC,WAAWC,OAAOH,MAAMC;IAC9BG,QAAQC,GAAG,CAACH;IACZ,IAAIA,SAASI,MAAM,GAAGZ,gBAAgB;QACpCU,QAAQG,IAAI,CAAC,CAAC,mBAAmB,EAAEC,KAAKC,KAAK,CAACP,SAASI,MAAM,GAAG,MAAM,WAAW,EAAEN,KAAKM,MAAM,CAAC,8EAA8E,CAAC;IAChL;AACF;AAEA,SAASH,OAAOH,IAAW,EAAEC,MAAwB;IACnD,IAAIA,WAAW,QAAQ,OAAOS,KAAKC,SAAS,CAACX,MAAM,MAAM;IACzD,IAAIA,KAAKM,MAAM,KAAK,GAAG,OAAO;IAE9B,MAAMM,UAAUC,OAAOC,IAAI,CAACd,IAAI,CAAC,EAAE;IACnC,MAAMe,QAAQf,KAAKgB,GAAG,CAAC,CAACC,MAAQL,QAAQI,GAAG,CAAC,CAACE,MAAQvB,KAAKsB,GAAG,CAACC,IAAI;IAClE,MAAMC,SAASP,QAAQI,GAAG,CAAC,CAACE,KAAKE,IAAMZ,KAAKa,GAAG,CAACH,IAAIZ,MAAM,KAAKS,MAAMC,GAAG,CAAC,CAACC,MAAQA,GAAG,CAACG,EAAE,CAACd,MAAM;IAE/F,MAAMgB,YAAY,CAACC,SAAqBA,OAAOP,GAAG,CAAC,CAACpB,OAAOwB,IAAMxB,MAAM4B,MAAM,CAACL,MAAM,CAACC,EAAE,GAAGK,IAAI,CAAC;IAE/F,OAAO;QAACH,UAAUV;QAAUO,OAAOH,GAAG,CAAC,CAACU,IAAM,IAAIC,MAAM,CAACD,IAAID,IAAI,CAAC;WAAUV,MAAMC,GAAG,CAACM;KAAW,CAACG,IAAI,CAAC;AACzG;AAEA,uEAAuE;AAEvE,OAAO,SAASG,UAAUC,MAAkH;IAC1I,MAAMC,QAAQ;QAAC,CAAC,MAAM,EAAED,OAAOE,IAAI,CAACC,KAAK,CAAC,EAAE,EAAExB,KAAKC,KAAK,CAACoB,OAAOE,IAAI,CAACE,KAAK,GAAG,MAAM,MAAM,CAAC;QAAE9B,OAAO0B,OAAOK,MAAM,EAAE;KAAS;IAC3H,IAAIL,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAC5B,MAAM,EAAEwB,MAAMM,IAAI,CAAC,CAAC,EAAE,EAAEP,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAC5B,MAAM,CAAC,aAAa,CAAC;IACvH,IAAIuB,OAAOQ,IAAI,CAAC/B,MAAM,GAAG,GAAGwB,MAAMM,IAAI,CAAC,0BAA0BjC,OAAO0B,OAAOQ,IAAI,EAAE;IACrFP,MAAMM,IAAI,CAAC,aAAajC,OAAO0B,OAAOS,MAAM,EAAE;IAC9C,OAAOR,MAAML,IAAI,CAAC;AACpB;AAEA,OAAO,SAASc,WAAWV,MAAkN;IAC3O,MAAMW,QAAQ;QAAC,GAAGX,OAAOY,IAAI,CAAC,IAAI,EAAEZ,OAAOa,MAAM,CAAC,QAAQ,CAAC;KAAC;IAC5D,KAAK,MAAM,CAACC,KAAK/C,MAAM,IAAIiB,OAAO+B,OAAO,CAACf,OAAOgB,WAAW,EAAGL,MAAMJ,IAAI,CAAC,CAAC,EAAE,EAAEO,IAAI,EAAE,EAAE/C,OAAO;IAC9F,IAAIiC,OAAOiB,QAAQ,CAACxC,MAAM,GAAG,GAAG;QAC9BkC,MAAMJ,IAAI,CAAC,IAAI;QACf,KAAK,MAAMW,KAAKlB,OAAOiB,QAAQ,CAAEN,MAAMJ,IAAI,CAAC,CAAC,EAAE,EAAE,IAAIT,MAAM,CAACoB,EAAEC,KAAK,EAAY,CAAC,EAAED,EAAEE,OAAO,CAAC,IAAI,EAAEF,EAAEG,UAAU,CAAC,CAAC,EAAEH,EAAEI,QAAQ,CAAC,GAAG,EAAEJ,EAAEL,MAAM,CAAC,EAAE,CAAC;IAChJ;IACA,MAAMU,WAAW,CAACC,OAAeC,OAAiBC;QAChD,IAAIA,QAAQ,GAAGf,MAAMJ,IAAI,CAAC,GAAGiB,MAAM,EAAE,EAAEE,MAAM,GAAG,EAAED,MAAM7B,IAAI,CAAC,QAAQ8B,QAAQD,MAAMhD,MAAM,GAAG,CAAC,GAAG,EAAEiD,QAAQD,MAAMhD,MAAM,CAAC,KAAK,CAAC,GAAG,IAAI;IACtI;IACA,IAAIuB,OAAO2B,aAAa,GAAG3B,OAAO4B,eAAe,GAAG5B,OAAO6B,cAAc,GAAG,GAAGlB,MAAMJ,IAAI,CAAC;IAC1FgB,SAAS,aAAavB,OAAO8B,QAAQ,EAAE9B,OAAO2B,aAAa;IAC3DJ,SAAS,cAAcvB,OAAO+B,UAAU,EAAE/B,OAAO4B,eAAe;IAChEL,SAAS,aAAavB,OAAOgC,SAAS,EAAEhC,OAAO6B,cAAc;IAC7D,OAAOlB,MAAMf,IAAI,CAAC;AACpB"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/output.ts"],"sourcesContent":["// rows -> table (default) | json\n\nexport type Row = Record<string, unknown>;\n\n// Warned about on stderr (keeps --format json stdout machine-readable), never truncated.\nconst OVERSIZE_BYTES = 50_000;\n\n// Flatten embedded newlines so they can't break table column alignment; JSON output is left alone.\nfunction cell(value: unknown): string {\n  return String(value ?? '').replace(/\\s*\\n\\s*/g, ' ');\n}\n\nexport function printRows(rows: Row[], format: 'table' | 'json'): void {\n  const rendered = renderRows(rows, format);\n  console.log(rendered);\n  if (rendered.length > OVERSIZE_BYTES) {\n    console.warn(`warning: result is ${Math.round(rendered.length / 1000)} KB across ${rows.length} row(s); consider a LIMIT, fewer columns, or snippet() instead of whole values`);\n  }\n}\n\n// Table output is for reading; a row wider than the terminal wraps into an unreadable\n// block (a `find` header measured 330 chars once `summary` is present, which drove one\n// user to pipe --format json through a script for every interactive query). Widest\n// columns give up space first, down to a floor, and only as far as the terminal needs.\n// --format json is never truncated: it is the machine-readable surface.\nconst MIN_COLUMN = 12;\nconst TRUNCATED = '…';\n\nfunction fitWidths(widths: number[], limit: number): number[] {\n  const gaps = (widths.length - 1) * 2;\n  const fitted = [...widths];\n  for (;;) {\n    const total = fitted.reduce((a, b) => a + b, 0) + gaps;\n    if (total <= limit) return fitted;\n    const widest = fitted.indexOf(Math.max(...fitted));\n    if (fitted[widest] <= MIN_COLUMN) return fitted;\n    fitted[widest] = Math.max(MIN_COLUMN, fitted[widest] - Math.max(1, total - limit));\n  }\n}\n\nfunction renderRows(rows: Row[], format: 'table' | 'json', width = process.stdout.columns): string {\n  if (format === 'json') return JSON.stringify(rows, null, 2);\n  if (rows.length === 0) return '(0 rows)';\n\n  const columns = Object.keys(rows[0]);\n  const cells = rows.map((row) => columns.map((col) => cell(row[col])));\n  const natural = columns.map((col, i) => Math.max(col.length, ...cells.map((row) => row[i].length)));\n  const widths = width && width > MIN_COLUMN ? fitWidths(natural, width) : natural;\n\n  const clip = (value: string, w: number) => (value.length <= w ? value.padEnd(w) : `${value.slice(0, w - 1)}${TRUNCATED}`);\n  const formatRow = (values: string[]) =>\n    values\n      .map((value, i) => clip(value, widths[i]))\n      .join('  ')\n      .trimEnd();\n\n  return [formatRow(columns), widths.map((w) => '-'.repeat(w)).join('  '), ...cells.map(formatRow)].join('\\n');\n}\n\n// Text renderers for the layer verbs; cli.ts prints what these return.\n\nfunction featuresLine(features: { on: string[]; off: string[] }): string {\n  const off = features.off.length > 0 ? ` · off: ${features.off.map((name) => `${name} (features.${name})`).join(', ')}` : '';\n  return `features: ${features.on.join(', ')}${off}`;\n}\n\nexport function renderMap(result: { docs: { count: number; bytes: number }; fields: Row[]; fieldsTotal: number; features: { on: string[]; off: string[] }; hubs: Row[]; recent: Row[] }): string {\n  const parts = [`docs: ${result.docs.count} (${Math.round(result.docs.bytes / 1024)} KB)`, `${featuresLine(result.features)}\\n`, renderRows(result.fields, 'table')];\n  if (result.fieldsTotal > result.fields.length) parts.push(`(+${result.fieldsTotal - result.fields.length} more fields)`);\n  if (result.hubs.length > 0) parts.push('\\nhubs (by link rank):', renderRows(result.hubs, 'table'));\n  else if (result.features.off.includes('rank')) parts.push('\\nhubs: off (features.rank)');\n  parts.push('\\nrecent:', renderRows(result.recent, 'table'));\n  return parts.join('\\n');\n}\n\nexport function renderPeek(result: { path: string; tokens: number; frontmatter: Row; sections: Row[]; outbound: string[]; backlinks: string[]; unresolved: string[]; outboundTotal: number; backlinksTotal: number; unresolvedTotal: number; off: string[] }): string {\n  const lines = [`${result.path}  (~${result.tokens} tokens)`];\n  for (const [key, value] of Object.entries(result.frontmatter)) lines.push(`  ${key}: ${value}`);\n  if (result.sections.length > 0) {\n    lines.push('', 'sections:');\n    for (const s of result.sections) lines.push(`  ${'#'.repeat(s.level as number)} ${s.heading}  [L${s.start_line}-${s.end_line}, ~${s.tokens}t]`);\n  } else if (result.off.includes('sections')) lines.push('', 'sections: off (features.sections)');\n  if (result.off.includes('links')) {\n    lines.push('', 'links: off (features.links)');\n    return lines.join('\\n');\n  }\n  const linkLine = (label: string, shown: string[], total: number) => {\n    if (total > 0) lines.push(`${label} (${total}): ${shown.join(', ')}${total > shown.length ? `, +${total - shown.length} more` : ''}`);\n  };\n  if (result.outboundTotal + result.unresolvedTotal + result.backlinksTotal > 0) lines.push('');\n  linkLine('links out', result.outbound, result.outboundTotal);\n  linkLine('unresolved', result.unresolved, result.unresolvedTotal);\n  linkLine('backlinks', result.backlinks, result.backlinksTotal);\n  return lines.join('\\n');\n}\n"],"names":["OVERSIZE_BYTES","cell","value","String","replace","printRows","rows","format","rendered","renderRows","console","log","length","warn","Math","round","MIN_COLUMN","TRUNCATED","fitWidths","widths","limit","gaps","fitted","total","reduce","a","b","widest","indexOf","max","width","process","stdout","columns","JSON","stringify","Object","keys","cells","map","row","col","natural","i","clip","w","padEnd","slice","formatRow","values","join","trimEnd","repeat","featuresLine","features","off","name","on","renderMap","result","parts","docs","count","bytes","fields","fieldsTotal","push","hubs","includes","recent","renderPeek","lines","path","tokens","key","entries","frontmatter","sections","s","level","heading","start_line","end_line","linkLine","label","shown","outboundTotal","unresolvedTotal","backlinksTotal","outbound","unresolved","backlinks"],"mappings":"AAAA,iCAAiC;AAIjC,yFAAyF;AACzF,MAAMA,iBAAiB;AAEvB,mGAAmG;AACnG,SAASC,KAAKC,KAAc;IAC1B,OAAOC,OAAOD,kBAAAA,mBAAAA,QAAS,IAAIE,OAAO,CAAC,aAAa;AAClD;AAEA,OAAO,SAASC,UAAUC,IAAW,EAAEC,MAAwB;IAC7D,MAAMC,WAAWC,WAAWH,MAAMC;IAClCG,QAAQC,GAAG,CAACH;IACZ,IAAIA,SAASI,MAAM,GAAGZ,gBAAgB;QACpCU,QAAQG,IAAI,CAAC,CAAC,mBAAmB,EAAEC,KAAKC,KAAK,CAACP,SAASI,MAAM,GAAG,MAAM,WAAW,EAAEN,KAAKM,MAAM,CAAC,8EAA8E,CAAC;IAChL;AACF;AAEA,sFAAsF;AACtF,uFAAuF;AACvF,mFAAmF;AACnF,uFAAuF;AACvF,wEAAwE;AACxE,MAAMI,aAAa;AACnB,MAAMC,YAAY;AAElB,SAASC,UAAUC,MAAgB,EAAEC,KAAa;IAChD,MAAMC,OAAO,AAACF,CAAAA,OAAOP,MAAM,GAAG,CAAA,IAAK;IACnC,MAAMU,SAAS;WAAIH;KAAO;IAC1B,OAAS;QACP,MAAMI,QAAQD,OAAOE,MAAM,CAAC,CAACC,GAAGC,IAAMD,IAAIC,GAAG,KAAKL;QAClD,IAAIE,SAASH,OAAO,OAAOE;QAC3B,MAAMK,SAASL,OAAOM,OAAO,CAACd,KAAKe,GAAG,IAAIP;QAC1C,IAAIA,MAAM,CAACK,OAAO,IAAIX,YAAY,OAAOM;QACzCA,MAAM,CAACK,OAAO,GAAGb,KAAKe,GAAG,CAACb,YAAYM,MAAM,CAACK,OAAO,GAAGb,KAAKe,GAAG,CAAC,GAAGN,QAAQH;IAC7E;AACF;AAEA,SAASX,WAAWH,IAAW,EAAEC,MAAwB,EAAEuB,QAAQC,QAAQC,MAAM,CAACC,OAAO;IACvF,IAAI1B,WAAW,QAAQ,OAAO2B,KAAKC,SAAS,CAAC7B,MAAM,MAAM;IACzD,IAAIA,KAAKM,MAAM,KAAK,GAAG,OAAO;IAE9B,MAAMqB,UAAUG,OAAOC,IAAI,CAAC/B,IAAI,CAAC,EAAE;IACnC,MAAMgC,QAAQhC,KAAKiC,GAAG,CAAC,CAACC,MAAQP,QAAQM,GAAG,CAAC,CAACE,MAAQxC,KAAKuC,GAAG,CAACC,IAAI;IAClE,MAAMC,UAAUT,QAAQM,GAAG,CAAC,CAACE,KAAKE,IAAM7B,KAAKe,GAAG,CAACY,IAAI7B,MAAM,KAAK0B,MAAMC,GAAG,CAAC,CAACC,MAAQA,GAAG,CAACG,EAAE,CAAC/B,MAAM;IAChG,MAAMO,SAASW,SAASA,QAAQd,aAAaE,UAAUwB,SAASZ,SAASY;IAEzE,MAAME,OAAO,CAAC1C,OAAe2C,IAAe3C,MAAMU,MAAM,IAAIiC,IAAI3C,MAAM4C,MAAM,CAACD,KAAK,GAAG3C,MAAM6C,KAAK,CAAC,GAAGF,IAAI,KAAK5B,WAAW;IACxH,MAAM+B,YAAY,CAACC,SACjBA,OACGV,GAAG,CAAC,CAACrC,OAAOyC,IAAMC,KAAK1C,OAAOiB,MAAM,CAACwB,EAAE,GACvCO,IAAI,CAAC,MACLC,OAAO;IAEZ,OAAO;QAACH,UAAUf;QAAUd,OAAOoB,GAAG,CAAC,CAACM,IAAM,IAAIO,MAAM,CAACP,IAAIK,IAAI,CAAC;WAAUZ,MAAMC,GAAG,CAACS;KAAW,CAACE,IAAI,CAAC;AACzG;AAEA,uEAAuE;AAEvE,SAASG,aAAaC,QAAyC;IAC7D,MAAMC,MAAMD,SAASC,GAAG,CAAC3C,MAAM,GAAG,IAAI,CAAC,QAAQ,EAAE0C,SAASC,GAAG,CAAChB,GAAG,CAAC,CAACiB,OAAS,GAAGA,KAAK,WAAW,EAAEA,KAAK,CAAC,CAAC,EAAEN,IAAI,CAAC,OAAO,GAAG;IACzH,OAAO,CAAC,UAAU,EAAEI,SAASG,EAAE,CAACP,IAAI,CAAC,QAAQK,KAAK;AACpD;AAEA,OAAO,SAASG,UAAUC,MAA6J;IACrL,MAAMC,QAAQ;QAAC,CAAC,MAAM,EAAED,OAAOE,IAAI,CAACC,KAAK,CAAC,EAAE,EAAEhD,KAAKC,KAAK,CAAC4C,OAAOE,IAAI,CAACE,KAAK,GAAG,MAAM,IAAI,CAAC;QAAE,GAAGV,aAAaM,OAAOL,QAAQ,EAAE,EAAE,CAAC;QAAE7C,WAAWkD,OAAOK,MAAM,EAAE;KAAS;IACnK,IAAIL,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAACpD,MAAM,EAAEgD,MAAMM,IAAI,CAAC,CAAC,EAAE,EAAEP,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAACpD,MAAM,CAAC,aAAa,CAAC;IACvH,IAAI+C,OAAOQ,IAAI,CAACvD,MAAM,GAAG,GAAGgD,MAAMM,IAAI,CAAC,0BAA0BzD,WAAWkD,OAAOQ,IAAI,EAAE;SACpF,IAAIR,OAAOL,QAAQ,CAACC,GAAG,CAACa,QAAQ,CAAC,SAASR,MAAMM,IAAI,CAAC;IAC1DN,MAAMM,IAAI,CAAC,aAAazD,WAAWkD,OAAOU,MAAM,EAAE;IAClD,OAAOT,MAAMV,IAAI,CAAC;AACpB;AAEA,OAAO,SAASoB,WAAWX,MAAiO;IAC1P,MAAMY,QAAQ;QAAC,GAAGZ,OAAOa,IAAI,CAAC,IAAI,EAAEb,OAAOc,MAAM,CAAC,QAAQ,CAAC;KAAC;IAC5D,KAAK,MAAM,CAACC,KAAKxE,MAAM,IAAIkC,OAAOuC,OAAO,CAAChB,OAAOiB,WAAW,EAAGL,MAAML,IAAI,CAAC,CAAC,EAAE,EAAEQ,IAAI,EAAE,EAAExE,OAAO;IAC9F,IAAIyD,OAAOkB,QAAQ,CAACjE,MAAM,GAAG,GAAG;QAC9B2D,MAAML,IAAI,CAAC,IAAI;QACf,KAAK,MAAMY,KAAKnB,OAAOkB,QAAQ,CAAEN,MAAML,IAAI,CAAC,CAAC,EAAE,EAAE,IAAId,MAAM,CAAC0B,EAAEC,KAAK,EAAY,CAAC,EAAED,EAAEE,OAAO,CAAC,IAAI,EAAEF,EAAEG,UAAU,CAAC,CAAC,EAAEH,EAAEI,QAAQ,CAAC,GAAG,EAAEJ,EAAEL,MAAM,CAAC,EAAE,CAAC;IAChJ,OAAO,IAAId,OAAOJ,GAAG,CAACa,QAAQ,CAAC,aAAaG,MAAML,IAAI,CAAC,IAAI;IAC3D,IAAIP,OAAOJ,GAAG,CAACa,QAAQ,CAAC,UAAU;QAChCG,MAAML,IAAI,CAAC,IAAI;QACf,OAAOK,MAAMrB,IAAI,CAAC;IACpB;IACA,MAAMiC,WAAW,CAACC,OAAeC,OAAiB9D;QAChD,IAAIA,QAAQ,GAAGgD,MAAML,IAAI,CAAC,GAAGkB,MAAM,EAAE,EAAE7D,MAAM,GAAG,EAAE8D,MAAMnC,IAAI,CAAC,QAAQ3B,QAAQ8D,MAAMzE,MAAM,GAAG,CAAC,GAAG,EAAEW,QAAQ8D,MAAMzE,MAAM,CAAC,KAAK,CAAC,GAAG,IAAI;IACtI;IACA,IAAI+C,OAAO2B,aAAa,GAAG3B,OAAO4B,eAAe,GAAG5B,OAAO6B,cAAc,GAAG,GAAGjB,MAAML,IAAI,CAAC;IAC1FiB,SAAS,aAAaxB,OAAO8B,QAAQ,EAAE9B,OAAO2B,aAAa;IAC3DH,SAAS,cAAcxB,OAAO+B,UAAU,EAAE/B,OAAO4B,eAAe;IAChEJ,SAAS,aAAaxB,OAAOgC,SAAS,EAAEhC,OAAO6B,cAAc;IAC7D,OAAOjB,MAAMrB,IAAI,CAAC;AACpB"}

package/dist/esm/scan.d.ts

@@ -11,7 +11,7 @@ export interface ParsedDoc {
     relPath: string;
     mtimeMs: number;
     size: number;
-    data: Record<string, string | number | null>;
+    data: Record<string, string | number | bigint | null>;
     search: {
         title: string;
         summary: string;

package/dist/esm/scan.js

@@ -41,10 +41,16 @@ export function listFiles(cfg, baseDir) {
         };
     });
 }
+// SQLite storage classes follow the YAML scalar: strings -> TEXT, booleans and whole
+// numbers -> INTEGER, fractions -> REAL, everything else (lists, maps) -> JSON TEXT.
+// Booleans have no SQLite type of their own; 1/0 is the convention, so `WHERE flag = 1`
+// matches and `WHERE flag = 'true'` cannot -- `map` prints the observed type per field so
+// that mismatch is visible rather than a silent zero-row answer.
 function mapValue(value) {
     if (value === null || value === undefined) return null;
-    if (typeof value === 'boolean') return value ? 1 : 0;
-    if (typeof value === 'string' || typeof value === 'number') return value;
+    if (typeof value === 'boolean') return BigInt(value ? 1 : 0);
+    if (typeof value === 'number') return Number.isSafeInteger(value) ? BigInt(value) : value;
+    if (typeof value === 'string') return value;
     return JSON.stringify(value);
 }
 // The delimiter split is all this package used gray-matter for.
@@ -99,24 +105,25 @@ export function parseFile(file, extractors = []) {
         }
         mapped[key] = mapValue(data[key]);
     }
+    // title/summary are plain YAML strings -- whitespace-collapse only;
+    // the prose gets the full markdown strip.
+    const search = {
+        title: normalizeText(data.title),
+        summary: normalizeText(data.summary),
+        text: stripText(content)
+    };
     return {
         doc: {
             relPath: file.relPath,
             mtimeMs: file.mtimeMs,
             size: file.size,
             data: mapped,
-            search: {
-                // title/summary are plain YAML strings -- whitespace-collapse only;
-                // the prose gets the full markdown strip.
-                title: normalizeText(data.title),
-                summary: normalizeText(data.summary),
-                text: stripText(content)
-            },
+            search,
             extracted: Object.fromEntries(extractors.filter((f)=>f.extract).map((f)=>{
                 var _f_extract;
                 return [
                     f.name,
-                    (_f_extract = f.extract) === null || _f_extract === void 0 ? void 0 : _f_extract.call(f, raw, content)
+                    (_f_extract = f.extract) === null || _f_extract === void 0 ? void 0 : _f_extract.call(f, raw, content, search)
                 ];
             }))
         },

package/dist/esm/scan.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/scan.ts"],"sourcesContent":["import { readFileSync, statSync } from 'node:fs';\nimport { join } from 'node:path';\nimport fastGlob from 'fast-glob';\nimport removeMarkdown from 'remove-markdown';\nimport { parseDocument } from 'yaml';\nimport type { Config } from './config.ts';\nimport type { Feature } from './features/types.ts';\n\n// Filesystem -> rows. Pure data in, data + warnings out; db.ts does the SQL.\n\n// Reserved: colliding with these would clash with the `frontmatter` table's own columns or the other tables (`content`, `links`, `sections`).\nconst RESERVED_COLUMNS = new Set(['path', '_mtime', '_size', '_rank', 'content', 'links', 'sections']);\n\nfunction normalizeText(value: unknown): string {\n  if (value === null || value === undefined) return '';\n  return String(value).replace(/\\s+/g, ' ').trim();\n}\n\n// Markdown is stripped at index time so snippets read as clean prose (also improves matching, e.g. `**bold**` indexes as `bold`).\n// remove-markdown doesn't cover Obsidian wikilinks or table scaffolding, so those are handled first/after.\nfunction stripText(value: string): string {\n  const withoutWikilinks = value.replace(/\\[\\[([^\\]|]+)\\|([^\\]]+)\\]\\]/g, '$2').replace(/\\[\\[([^\\]]+)\\]\\]/g, '$1');\n  const withoutMarkdown = removeMarkdown(withoutWikilinks);\n  const withoutTables = withoutMarkdown.replace(/^\\s*\\|?[-\\s|:]+\\|\\s*$/gm, '').replace(/\\|/g, ' ');\n  return normalizeText(withoutTables);\n}\n\nexport interface FileStat {\n  relPath: string;\n  absPath: string;\n  mtimeMs: number;\n  size: number;\n}\n\nexport function listFiles(cfg: Config, baseDir: string): FileStat[] {\n  const relPaths = fastGlob.sync(cfg.scan.include, { cwd: baseDir }).sort();\n  return relPaths.map((relPath) => {\n    const absPath = join(baseDir, relPath);\n    const st = statSync(absPath);\n    return { relPath, absPath, mtimeMs: st.mtimeMs, size: st.size };\n  });\n}\n\nexport interface ParsedDoc {\n  relPath: string;\n  mtimeMs: number;\n  size: number;\n  data: Record<string, string | number | null>;\n  // title/summary are duplicated from frontmatter so bm25() can weight them above the body text.\n  search: { title: string; summary: string; text: string };\n  // Per-feature extraction results, keyed by feature name; features store them at reconcile.\n  extracted: Record<string, unknown>;\n}\n\nfunction mapValue(value: unknown): string | number | null {\n  if (value === null || value === undefined) return null;\n  if (typeof value === 'boolean') return value ? 1 : 0;\n  if (typeof value === 'string' || typeof value === 'number') return value;\n  return JSON.stringify(value);\n}\n\n// The delimiter split is all this package used gray-matter for.\nfunction splitFrontmatter(raw: string): { fm: string | null; body: string } {\n  const open = raw.match(/^---\\r?\\n/);\n  if (!open) return { fm: null, body: raw };\n  const rest = raw.slice(open[0].length);\n  const close = rest.match(/^---\\r?(\\n|$)/m);\n  if (!close || close.index === undefined) return { fm: null, body: raw };\n  return { fm: rest.slice(0, close.index), body: rest.slice(close.index + close[0].length) };\n}\n\n// Lenient by design: parseDocument collects syntax errors as data and still yields values,\n// so Obsidian-style frontmatter (e.g. an alias starting with @) survives with a warning.\nfunction parseFrontmatter(relPath: string, fm: string, warnings: string[]): Record<string, unknown> {\n  const doc = parseDocument(fm);\n  if (doc.errors.length > 0) {\n    warnings.push(`warning: ${relPath} frontmatter has ${doc.errors.length} syntax error(s) (${doc.errors[0].message.split('\\n')[0]}); parsed leniently`);\n  }\n  let data: unknown;\n  try {\n    data = doc.toJS();\n  } catch (err) {\n    warnings.push(`warning: ${relPath} has unparseable frontmatter (${(err as Error).message.split('\\n')[0]}); indexing without it`);\n    return {};\n  }\n  if (data === null || data === undefined) return {};\n  if (typeof data !== 'object' || Array.isArray(data)) {\n    warnings.push(`warning: ${relPath} frontmatter is not a key-value mapping; ignoring it`);\n    return {};\n  }\n  return data as Record<string, unknown>;\n}\n\nexport function parseFile(file: FileStat, extractors: Feature[] = []): { doc: ParsedDoc; warnings: string[] } {\n  const raw = readFileSync(file.absPath, 'utf8');\n  const warnings: string[] = [];\n\n  const { fm, body: content } = splitFrontmatter(raw);\n  const data = fm === null ? {} : parseFrontmatter(file.relPath, fm, warnings);\n  const mapped: Record<string, string | number | null> = {};\n\n  for (const key of Object.keys(data)) {\n    if (RESERVED_COLUMNS.has(key)) {\n      warnings.push(`warning: ${file.relPath} has a frontmatter key named \"${key}\", which is reserved; ignoring it`);\n      continue;\n    }\n    mapped[key] = mapValue(data[key]);\n  }\n\n  return {\n    doc: {\n      relPath: file.relPath,\n      mtimeMs: file.mtimeMs,\n      size: file.size,\n      data: mapped,\n      search: {\n        // title/summary are plain YAML strings -- whitespace-collapse only;\n        // the prose gets the full markdown strip.\n        title: normalizeText(data.title),\n        summary: normalizeText(data.summary),\n        text: stripText(content),\n      },\n      extracted: Object.fromEntries(extractors.filter((f) => f.extract).map((f) => [f.name, f.extract?.(raw, content)])),\n    },\n    warnings,\n  };\n}\n"],"names":["readFileSync","statSync","join","fastGlob","removeMarkdown","parseDocument","RESERVED_COLUMNS","Set","normalizeText","value","undefined","String","replace","trim","stripText","withoutWikilinks","withoutMarkdown","withoutTables","listFiles","cfg","baseDir","relPaths","sync","scan","include","cwd","sort","map","relPath","absPath","st","mtimeMs","size","mapValue","JSON","stringify","splitFrontmatter","raw","open","match","fm","body","rest","slice","length","close","index","parseFrontmatter","warnings","doc","errors","push","message","split","data","toJS","err","Array","isArray","parseFile","file","extractors","content","mapped","key","Object","keys","has","search","title","summary","text","extracted","fromEntries","filter","f","extract","name"],"mappings":"AAAA,SAASA,YAAY,EAAEC,QAAQ,QAAQ,UAAU;AACjD,SAASC,IAAI,QAAQ,YAAY;AACjC,OAAOC,cAAc,YAAY;AACjC,OAAOC,oBAAoB,kBAAkB;AAC7C,SAASC,aAAa,QAAQ,OAAO;AAIrC,6EAA6E;AAE7E,8IAA8I;AAC9I,MAAMC,mBAAmB,IAAIC,IAAI;IAAC;IAAQ;IAAU;IAAS;IAAS;IAAW;IAAS;CAAW;AAErG,SAASC,cAAcC,KAAc;IACnC,IAAIA,UAAU,QAAQA,UAAUC,WAAW,OAAO;IAClD,OAAOC,OAAOF,OAAOG,OAAO,CAAC,QAAQ,KAAKC,IAAI;AAChD;AAEA,kIAAkI;AAClI,2GAA2G;AAC3G,SAASC,UAAUL,KAAa;IAC9B,MAAMM,mBAAmBN,MAAMG,OAAO,CAAC,gCAAgC,MAAMA,OAAO,CAAC,qBAAqB;IAC1G,MAAMI,kBAAkBZ,eAAeW;IACvC,MAAME,gBAAgBD,gBAAgBJ,OAAO,CAAC,2BAA2B,IAAIA,OAAO,CAAC,OAAO;IAC5F,OAAOJ,cAAcS;AACvB;AASA,OAAO,SAASC,UAAUC,GAAW,EAAEC,OAAe;IACpD,MAAMC,WAAWlB,SAASmB,IAAI,CAACH,IAAII,IAAI,CAACC,OAAO,EAAE;QAAEC,KAAKL;IAAQ,GAAGM,IAAI;IACvE,OAAOL,SAASM,GAAG,CAAC,CAACC;QACnB,MAAMC,UAAU3B,KAAKkB,SAASQ;QAC9B,MAAME,KAAK7B,SAAS4B;QACpB,OAAO;YAAED;YAASC;YAASE,SAASD,GAAGC,OAAO;YAAEC,MAAMF,GAAGE,IAAI;QAAC;IAChE;AACF;AAaA,SAASC,SAASxB,KAAc;IAC9B,IAAIA,UAAU,QAAQA,UAAUC,WAAW,OAAO;IAClD,IAAI,OAAOD,UAAU,WAAW,OAAOA,QAAQ,IAAI;IACnD,IAAI,OAAOA,UAAU,YAAY,OAAOA,UAAU,UAAU,OAAOA;IACnE,OAAOyB,KAAKC,SAAS,CAAC1B;AACxB;AAEA,gEAAgE;AAChE,SAAS2B,iBAAiBC,GAAW;IACnC,MAAMC,OAAOD,IAAIE,KAAK,CAAC;IACvB,IAAI,CAACD,MAAM,OAAO;QAAEE,IAAI;QAAMC,MAAMJ;IAAI;IACxC,MAAMK,OAAOL,IAAIM,KAAK,CAACL,IAAI,CAAC,EAAE,CAACM,MAAM;IACrC,MAAMC,QAAQH,KAAKH,KAAK,CAAC;IACzB,IAAI,CAACM,SAASA,MAAMC,KAAK,KAAKpC,WAAW,OAAO;QAAE8B,IAAI;QAAMC,MAAMJ;IAAI;IACtE,OAAO;QAAEG,IAAIE,KAAKC,KAAK,CAAC,GAAGE,MAAMC,KAAK;QAAGL,MAAMC,KAAKC,KAAK,CAACE,MAAMC,KAAK,GAAGD,KAAK,CAAC,EAAE,CAACD,MAAM;IAAE;AAC3F;AAEA,2FAA2F;AAC3F,yFAAyF;AACzF,SAASG,iBAAiBnB,OAAe,EAAEY,EAAU,EAAEQ,QAAkB;IACvE,MAAMC,MAAM5C,cAAcmC;IAC1B,IAAIS,IAAIC,MAAM,CAACN,MAAM,GAAG,GAAG;QACzBI,SAASG,IAAI,CAAC,CAAC,SAAS,EAAEvB,QAAQ,iBAAiB,EAAEqB,IAAIC,MAAM,CAACN,MAAM,CAAC,kBAAkB,EAAEK,IAAIC,MAAM,CAAC,EAAE,CAACE,OAAO,CAACC,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC,mBAAmB,CAAC;IACtJ;IACA,IAAIC;IACJ,IAAI;QACFA,OAAOL,IAAIM,IAAI;IACjB,EAAE,OAAOC,KAAK;QACZR,SAASG,IAAI,CAAC,CAAC,SAAS,EAAEvB,QAAQ,8BAA8B,EAAE,AAAC4B,IAAcJ,OAAO,CAACC,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC,sBAAsB,CAAC;QAC/H,OAAO,CAAC;IACV;IACA,IAAIC,SAAS,QAAQA,SAAS5C,WAAW,OAAO,CAAC;IACjD,IAAI,OAAO4C,SAAS,YAAYG,MAAMC,OAAO,CAACJ,OAAO;QACnDN,SAASG,IAAI,CAAC,CAAC,SAAS,EAAEvB,QAAQ,oDAAoD,CAAC;QACvF,OAAO,CAAC;IACV;IACA,OAAO0B;AACT;AAEA,OAAO,SAASK,UAAUC,IAAc,EAAEC,aAAwB,EAAE;IAClE,MAAMxB,MAAMrC,aAAa4D,KAAK/B,OAAO,EAAE;IACvC,MAAMmB,WAAqB,EAAE;IAE7B,MAAM,EAAER,EAAE,EAAEC,MAAMqB,OAAO,EAAE,GAAG1B,iBAAiBC;IAC/C,MAAMiB,OAAOd,OAAO,OAAO,CAAC,IAAIO,iBAAiBa,KAAKhC,OAAO,EAAEY,IAAIQ;IACnE,MAAMe,SAAiD,CAAC;IAExD,KAAK,MAAMC,OAAOC,OAAOC,IAAI,CAACZ,MAAO;QACnC,IAAIhD,iBAAiB6D,GAAG,CAACH,MAAM;YAC7BhB,SAASG,IAAI,CAAC,CAAC,SAAS,EAAES,KAAKhC,OAAO,CAAC,8BAA8B,EAAEoC,IAAI,iCAAiC,CAAC;YAC7G;QACF;QACAD,MAAM,CAACC,IAAI,GAAG/B,SAASqB,IAAI,CAACU,IAAI;IAClC;IAEA,OAAO;QACLf,KAAK;YACHrB,SAASgC,KAAKhC,OAAO;YACrBG,SAAS6B,KAAK7B,OAAO;YACrBC,MAAM4B,KAAK5B,IAAI;YACfsB,MAAMS;YACNK,QAAQ;gBACN,oEAAoE;gBACpE,0CAA0C;gBAC1CC,OAAO7D,cAAc8C,KAAKe,KAAK;gBAC/BC,SAAS9D,cAAc8C,KAAKgB,OAAO;gBACnCC,MAAMzD,UAAUgD;YAClB;YACAU,WAAWP,OAAOQ,WAAW,CAACZ,WAAWa,MAAM,CAAC,CAACC,IAAMA,EAAEC,OAAO,EAAEjD,GAAG,CAAC,CAACgD;oBAAeA;uBAAT;oBAACA,EAAEE,IAAI;qBAAEF,aAAAA,EAAEC,OAAO,cAATD,iCAAAA,gBAAAA,GAAYtC,KAAKyB;iBAAS;;QAClH;QACAd;IACF;AACF"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/scan.ts"],"sourcesContent":["import { readFileSync, statSync } from 'node:fs';\nimport { join } from 'node:path';\nimport fastGlob from 'fast-glob';\nimport removeMarkdown from 'remove-markdown';\nimport { parseDocument } from 'yaml';\nimport type { Config } from './config.ts';\nimport type { Feature } from './features/types.ts';\n\n// Filesystem -> rows. Pure data in, data + warnings out; db.ts does the SQL.\n\n// Reserved: colliding with these would clash with the `frontmatter` table's own columns or the other tables (`content`, `links`, `sections`).\nconst RESERVED_COLUMNS = new Set(['path', '_mtime', '_size', '_rank', 'content', 'links', 'sections']);\n\nfunction normalizeText(value: unknown): string {\n  if (value === null || value === undefined) return '';\n  return String(value).replace(/\\s+/g, ' ').trim();\n}\n\n// Markdown is stripped at index time so snippets read as clean prose (also improves matching, e.g. `**bold**` indexes as `bold`).\n// remove-markdown doesn't cover Obsidian wikilinks or table scaffolding, so those are handled first/after.\nfunction stripText(value: string): string {\n  const withoutWikilinks = value.replace(/\\[\\[([^\\]|]+)\\|([^\\]]+)\\]\\]/g, '$2').replace(/\\[\\[([^\\]]+)\\]\\]/g, '$1');\n  const withoutMarkdown = removeMarkdown(withoutWikilinks);\n  const withoutTables = withoutMarkdown.replace(/^\\s*\\|?[-\\s|:]+\\|\\s*$/gm, '').replace(/\\|/g, ' ');\n  return normalizeText(withoutTables);\n}\n\nexport interface FileStat {\n  relPath: string;\n  absPath: string;\n  mtimeMs: number;\n  size: number;\n}\n\nexport function listFiles(cfg: Config, baseDir: string): FileStat[] {\n  const relPaths = fastGlob.sync(cfg.scan.include, { cwd: baseDir }).sort();\n  return relPaths.map((relPath) => {\n    const absPath = join(baseDir, relPath);\n    const st = statSync(absPath);\n    return { relPath, absPath, mtimeMs: st.mtimeMs, size: st.size };\n  });\n}\n\nexport interface ParsedDoc {\n  relPath: string;\n  mtimeMs: number;\n  size: number;\n  data: Record<string, string | number | bigint | null>;\n  // title/summary are duplicated from frontmatter so bm25() can weight them above the body text.\n  search: { title: string; summary: string; text: string };\n  // Per-feature extraction results, keyed by feature name; features store them at reconcile.\n  extracted: Record<string, unknown>;\n}\n\n// SQLite storage classes follow the YAML scalar: strings -> TEXT, booleans and whole\n// numbers -> INTEGER, fractions -> REAL, everything else (lists, maps) -> JSON TEXT.\n// Booleans have no SQLite type of their own; 1/0 is the convention, so `WHERE flag = 1`\n// matches and `WHERE flag = 'true'` cannot -- `map` prints the observed type per field so\n// that mismatch is visible rather than a silent zero-row answer.\nfunction mapValue(value: unknown): string | number | bigint | null {\n  if (value === null || value === undefined) return null;\n  if (typeof value === 'boolean') return BigInt(value ? 1 : 0);\n  if (typeof value === 'number') return Number.isSafeInteger(value) ? BigInt(value) : value;\n  if (typeof value === 'string') return value;\n  return JSON.stringify(value);\n}\n\n// The delimiter split is all this package used gray-matter for.\nfunction splitFrontmatter(raw: string): { fm: string | null; body: string } {\n  const open = raw.match(/^---\\r?\\n/);\n  if (!open) return { fm: null, body: raw };\n  const rest = raw.slice(open[0].length);\n  const close = rest.match(/^---\\r?(\\n|$)/m);\n  if (!close || close.index === undefined) return { fm: null, body: raw };\n  return { fm: rest.slice(0, close.index), body: rest.slice(close.index + close[0].length) };\n}\n\n// Lenient by design: parseDocument collects syntax errors as data and still yields values,\n// so Obsidian-style frontmatter (e.g. an alias starting with @) survives with a warning.\nfunction parseFrontmatter(relPath: string, fm: string, warnings: string[]): Record<string, unknown> {\n  const doc = parseDocument(fm);\n  if (doc.errors.length > 0) {\n    warnings.push(`warning: ${relPath} frontmatter has ${doc.errors.length} syntax error(s) (${doc.errors[0].message.split('\\n')[0]}); parsed leniently`);\n  }\n  let data: unknown;\n  try {\n    data = doc.toJS();\n  } catch (err) {\n    warnings.push(`warning: ${relPath} has unparseable frontmatter (${(err as Error).message.split('\\n')[0]}); indexing without it`);\n    return {};\n  }\n  if (data === null || data === undefined) return {};\n  if (typeof data !== 'object' || Array.isArray(data)) {\n    warnings.push(`warning: ${relPath} frontmatter is not a key-value mapping; ignoring it`);\n    return {};\n  }\n  return data as Record<string, unknown>;\n}\n\nexport function parseFile(file: FileStat, extractors: Feature[] = []): { doc: ParsedDoc; warnings: string[] } {\n  const raw = readFileSync(file.absPath, 'utf8');\n  const warnings: string[] = [];\n\n  const { fm, body: content } = splitFrontmatter(raw);\n  const data = fm === null ? {} : parseFrontmatter(file.relPath, fm, warnings);\n  const mapped: Record<string, string | number | bigint | null> = {};\n\n  for (const key of Object.keys(data)) {\n    if (RESERVED_COLUMNS.has(key)) {\n      warnings.push(`warning: ${file.relPath} has a frontmatter key named \"${key}\", which is reserved; ignoring it`);\n      continue;\n    }\n    mapped[key] = mapValue(data[key]);\n  }\n\n  // title/summary are plain YAML strings -- whitespace-collapse only;\n  // the prose gets the full markdown strip.\n  const search = { title: normalizeText(data.title), summary: normalizeText(data.summary), text: stripText(content) };\n\n  return {\n    doc: {\n      relPath: file.relPath,\n      mtimeMs: file.mtimeMs,\n      size: file.size,\n      data: mapped,\n      search,\n      extracted: Object.fromEntries(extractors.filter((f) => f.extract).map((f) => [f.name, f.extract?.(raw, content, search)])),\n    },\n    warnings,\n  };\n}\n"],"names":["readFileSync","statSync","join","fastGlob","removeMarkdown","parseDocument","RESERVED_COLUMNS","Set","normalizeText","value","undefined","String","replace","trim","stripText","withoutWikilinks","withoutMarkdown","withoutTables","listFiles","cfg","baseDir","relPaths","sync","scan","include","cwd","sort","map","relPath","absPath","st","mtimeMs","size","mapValue","BigInt","Number","isSafeInteger","JSON","stringify","splitFrontmatter","raw","open","match","fm","body","rest","slice","length","close","index","parseFrontmatter","warnings","doc","errors","push","message","split","data","toJS","err","Array","isArray","parseFile","file","extractors","content","mapped","key","Object","keys","has","search","title","summary","text","extracted","fromEntries","filter","f","extract","name"],"mappings":"AAAA,SAASA,YAAY,EAAEC,QAAQ,QAAQ,UAAU;AACjD,SAASC,IAAI,QAAQ,YAAY;AACjC,OAAOC,cAAc,YAAY;AACjC,OAAOC,oBAAoB,kBAAkB;AAC7C,SAASC,aAAa,QAAQ,OAAO;AAIrC,6EAA6E;AAE7E,8IAA8I;AAC9I,MAAMC,mBAAmB,IAAIC,IAAI;IAAC;IAAQ;IAAU;IAAS;IAAS;IAAW;IAAS;CAAW;AAErG,SAASC,cAAcC,KAAc;IACnC,IAAIA,UAAU,QAAQA,UAAUC,WAAW,OAAO;IAClD,OAAOC,OAAOF,OAAOG,OAAO,CAAC,QAAQ,KAAKC,IAAI;AAChD;AAEA,kIAAkI;AAClI,2GAA2G;AAC3G,SAASC,UAAUL,KAAa;IAC9B,MAAMM,mBAAmBN,MAAMG,OAAO,CAAC,gCAAgC,MAAMA,OAAO,CAAC,qBAAqB;IAC1G,MAAMI,kBAAkBZ,eAAeW;IACvC,MAAME,gBAAgBD,gBAAgBJ,OAAO,CAAC,2BAA2B,IAAIA,OAAO,CAAC,OAAO;IAC5F,OAAOJ,cAAcS;AACvB;AASA,OAAO,SAASC,UAAUC,GAAW,EAAEC,OAAe;IACpD,MAAMC,WAAWlB,SAASmB,IAAI,CAACH,IAAII,IAAI,CAACC,OAAO,EAAE;QAAEC,KAAKL;IAAQ,GAAGM,IAAI;IACvE,OAAOL,SAASM,GAAG,CAAC,CAACC;QACnB,MAAMC,UAAU3B,KAAKkB,SAASQ;QAC9B,MAAME,KAAK7B,SAAS4B;QACpB,OAAO;YAAED;YAASC;YAASE,SAASD,GAAGC,OAAO;YAAEC,MAAMF,GAAGE,IAAI;QAAC;IAChE;AACF;AAaA,qFAAqF;AACrF,qFAAqF;AACrF,wFAAwF;AACxF,0FAA0F;AAC1F,iEAAiE;AACjE,SAASC,SAASxB,KAAc;IAC9B,IAAIA,UAAU,QAAQA,UAAUC,WAAW,OAAO;IAClD,IAAI,OAAOD,UAAU,WAAW,OAAOyB,OAAOzB,QAAQ,IAAI;IAC1D,IAAI,OAAOA,UAAU,UAAU,OAAO0B,OAAOC,aAAa,CAAC3B,SAASyB,OAAOzB,SAASA;IACpF,IAAI,OAAOA,UAAU,UAAU,OAAOA;IACtC,OAAO4B,KAAKC,SAAS,CAAC7B;AACxB;AAEA,gEAAgE;AAChE,SAAS8B,iBAAiBC,GAAW;IACnC,MAAMC,OAAOD,IAAIE,KAAK,CAAC;IACvB,IAAI,CAACD,MAAM,OAAO;QAAEE,IAAI;QAAMC,MAAMJ;IAAI;IACxC,MAAMK,OAAOL,IAAIM,KAAK,CAACL,IAAI,CAAC,EAAE,CAACM,MAAM;IACrC,MAAMC,QAAQH,KAAKH,KAAK,CAAC;IACzB,IAAI,CAACM,SAASA,MAAMC,KAAK,KAAKvC,WAAW,OAAO;QAAEiC,IAAI;QAAMC,MAAMJ;IAAI;IACtE,OAAO;QAAEG,IAAIE,KAAKC,KAAK,CAAC,GAAGE,MAAMC,KAAK;QAAGL,MAAMC,KAAKC,KAAK,CAACE,MAAMC,KAAK,GAAGD,KAAK,CAAC,EAAE,CAACD,MAAM;IAAE;AAC3F;AAEA,2FAA2F;AAC3F,yFAAyF;AACzF,SAASG,iBAAiBtB,OAAe,EAAEe,EAAU,EAAEQ,QAAkB;IACvE,MAAMC,MAAM/C,cAAcsC;IAC1B,IAAIS,IAAIC,MAAM,CAACN,MAAM,GAAG,GAAG;QACzBI,SAASG,IAAI,CAAC,CAAC,SAAS,EAAE1B,QAAQ,iBAAiB,EAAEwB,IAAIC,MAAM,CAACN,MAAM,CAAC,kBAAkB,EAAEK,IAAIC,MAAM,CAAC,EAAE,CAACE,OAAO,CAACC,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC,mBAAmB,CAAC;IACtJ;IACA,IAAIC;IACJ,IAAI;QACFA,OAAOL,IAAIM,IAAI;IACjB,EAAE,OAAOC,KAAK;QACZR,SAASG,IAAI,CAAC,CAAC,SAAS,EAAE1B,QAAQ,8BAA8B,EAAE,AAAC+B,IAAcJ,OAAO,CAACC,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC,sBAAsB,CAAC;QAC/H,OAAO,CAAC;IACV;IACA,IAAIC,SAAS,QAAQA,SAAS/C,WAAW,OAAO,CAAC;IACjD,IAAI,OAAO+C,SAAS,YAAYG,MAAMC,OAAO,CAACJ,OAAO;QACnDN,SAASG,IAAI,CAAC,CAAC,SAAS,EAAE1B,QAAQ,oDAAoD,CAAC;QACvF,OAAO,CAAC;IACV;IACA,OAAO6B;AACT;AAEA,OAAO,SAASK,UAAUC,IAAc,EAAEC,aAAwB,EAAE;IAClE,MAAMxB,MAAMxC,aAAa+D,KAAKlC,OAAO,EAAE;IACvC,MAAMsB,WAAqB,EAAE;IAE7B,MAAM,EAAER,EAAE,EAAEC,MAAMqB,OAAO,EAAE,GAAG1B,iBAAiBC;IAC/C,MAAMiB,OAAOd,OAAO,OAAO,CAAC,IAAIO,iBAAiBa,KAAKnC,OAAO,EAAEe,IAAIQ;IACnE,MAAMe,SAA0D,CAAC;IAEjE,KAAK,MAAMC,OAAOC,OAAOC,IAAI,CAACZ,MAAO;QACnC,IAAInD,iBAAiBgE,GAAG,CAACH,MAAM;YAC7BhB,SAASG,IAAI,CAAC,CAAC,SAAS,EAAES,KAAKnC,OAAO,CAAC,8BAA8B,EAAEuC,IAAI,iCAAiC,CAAC;YAC7G;QACF;QACAD,MAAM,CAACC,IAAI,GAAGlC,SAASwB,IAAI,CAACU,IAAI;IAClC;IAEA,oEAAoE;IACpE,0CAA0C;IAC1C,MAAMI,SAAS;QAAEC,OAAOhE,cAAciD,KAAKe,KAAK;QAAGC,SAASjE,cAAciD,KAAKgB,OAAO;QAAGC,MAAM5D,UAAUmD;IAAS;IAElH,OAAO;QACLb,KAAK;YACHxB,SAASmC,KAAKnC,OAAO;YACrBG,SAASgC,KAAKhC,OAAO;YACrBC,MAAM+B,KAAK/B,IAAI;YACfyB,MAAMS;YACNK;YACAI,WAAWP,OAAOQ,WAAW,CAACZ,WAAWa,MAAM,CAAC,CAACC,IAAMA,EAAEC,OAAO,EAAEpD,GAAG,CAAC,CAACmD;oBAAeA;uBAAT;oBAACA,EAAEE,IAAI;qBAAEF,aAAAA,EAAEC,OAAO,cAATD,iCAAAA,gBAAAA,GAAYtC,KAAKyB,SAASM;iBAAQ;;QAC1H;QACApB;IACF;AACF"}

package/dist/esm/verbs.d.ts

@@ -1,11 +1,12 @@
 import type { DatabaseSync } from 'node:sqlite';
-import type { Config } from './config.js';
+import type { Config, FeatureName } from './config.js';
 import type { Row } from './output.js';
 export interface FindOptions {
     k?: number;
     where?: string;
+    semantic?: boolean;
 }
-export declare function find(db: DatabaseSync, cfg: Config, terms: string, opts?: FindOptions): Row[];
+export declare function find(db: DatabaseSync, cfg: Config, terms: string, opts?: FindOptions): Promise<Row[]>;
 export interface TreeMap {
     docs: {
         count: number;
@@ -13,6 +14,10 @@ export interface TreeMap {
     };
     fields: Row[];
     fieldsTotal: number;
+    features: {
+        on: FeatureName[];
+        off: FeatureName[];
+    };
     hubs: Row[];
     recent: Row[];
 }
@@ -28,5 +33,6 @@ export interface Peek {
     outboundTotal: number;
     backlinksTotal: number;
     unresolvedTotal: number;
+    off: FeatureName[];
 }
 export declare function peek(db: DatabaseSync, cfg: Config, pathArg: string): Peek;

package/dist/esm/verbs.js

@@ -1,26 +1,62 @@
 import posix from 'node:path/posix';
-import { featureEnabled } from './config.js';
+import { featureEnabled, featureStates } from './config.js';
 import { SenseError } from './errors.js';
+import { semanticCandidates } from './features/embed.js';
 import { linkEdges } from './features/index.js';
 import { personalizedRank } from './graph.js';
 // The three layer verbs: mapTree (orient), find (locate), peek (structure).
 // Each returns data; cli.ts renders. All of them degrade when a feature is off.
 const WEIGHTED_BM25 = 'bm25(content, 10.0, 5.0, 1.0)';
 const RRF_K = 60;
+// FTS5 reads these as operators, so a bare term containing one is a syntax error or --
+// worse -- a column filter: `end-to-end` parses as a filter on column `to`, and SQLite
+// reports `no such column: to`, which is a true statement about the parse and a false
+// one about the input. Both field reports on 0.6.0 misdiagnosed that message.
+const FTS5_OPERATORS = /[-'"/.:^*()]/;
+function searchError(err, terms, scope) {
+    var _terms_match;
+    var _exec;
+    const message = err.message;
+    if (!/no such column|fts5: syntax error|malformed MATCH/.test(message)) return err;
+    const suspects = ((_terms_match = terms.match(/\S+/g)) !== null && _terms_match !== void 0 ? _terms_match : []).filter((t)=>!t.startsWith('"') && FTS5_OPERATORS.test(t));
+    // Blame the terms only when the failing token actually came from one -- a typo'd column
+    // in --where (or the tree's default scope) raises "no such column" through this same
+    // statement, and naming a term for it would state a false fact about the input.
+    const col = (_exec = /no such column: (\S+)/.exec(message)) === null || _exec === void 0 ? void 0 : _exec[1];
+    const fromTerms = col === undefined ? suspects.length > 0 : suspects.some((t)=>t.split(/[^\p{L}\p{N}]+/u).includes(col));
+    if (fromTerms && suspects.length > 0) {
+        return new SenseError('SEARCH_SYNTAX', `${message} -- the punctuation in ${suspects.map((t)=>`\`${t}\``).join(', ')} is FTS5 syntax, not literal text; search for it literally by double-quoting: '"${suspects[0]}"'. Searchable columns are title, summary, text.`);
+    }
+    if (col !== undefined && scope !== undefined) {
+        return new SenseError('SEARCH_SYNTAX', `${message} -- the where condition (${scope}) references it; frontmatter columns are listed by sense query "SELECT name FROM pragma_table_info('frontmatter')".`);
+    }
+    return new SenseError('SEARCH_SYNTAX', `${message} -- searchable columns are title, summary, text; frontmatter fields are queried with --where or sense query (list them with pragma_table_info('frontmatter')).`);
+}
 // Layer 1: BM25 + link-graph expansion, fused by reciprocal rank. `via` says which
 // signal produced each row so the agent knows what evidence it is trusting.
-export function find(db, cfg, terms, opts = {}) {
-    var _opts_k, _hits_get;
+// Semantic expansion is per-query opt-in: without opts.semantic the result is
+// byte-for-byte independent of the embed feature.
+export async function find(db, cfg, terms, opts = {}) {
+    var _opts_k, _opts_where, _hits_get, _chunkLines_get, _chunkSimilarity_get;
+    var _cfg_defaults_find, _cfg_defaults;
     const k = (_opts_k = opts.k) !== null && _opts_k !== void 0 ? _opts_k : 10;
     const fetch = Math.max(k * 3, 30);
     // Terms pass verbatim to FTS5 MATCH: bare words AND-join, operators are the caller's.
     // Invalid syntax propagates as an error, zero matches return zero -- no silent rewrites.
     // --where applies inside the candidate query (a post-filter over the top-N would drop
     // matches ranked past the pool) and again on the final select for link-derived rows.
-    const whereJoin = opts.where ? `JOIN frontmatter f ON f."path" = content.path` : '';
-    const whereCond = opts.where ? `AND (${opts.where})` : '';
+    // An explicit --where replaces the tree's declared default rather than ANDing with it,
+    // so a caller can always widen back to the whole tree.
+    const scope = (_opts_where = opts.where) !== null && _opts_where !== void 0 ? _opts_where : (_cfg_defaults = cfg.defaults) === null || _cfg_defaults === void 0 ? void 0 : (_cfg_defaults_find = _cfg_defaults.find) === null || _cfg_defaults_find === void 0 ? void 0 : _cfg_defaults_find.where;
+    const whereJoin = scope ? `JOIN frontmatter f ON f."path" = content.path` : '';
+    const whereCond = scope ? `AND (${scope})` : '';
     const matchSql = `SELECT content.path AS path, snippet(content, -1, '«', '»', '…', 10) AS hit FROM content ${whereJoin} WHERE content MATCH ? ${whereCond} ORDER BY ${WEIGHTED_BM25} LIMIT ${fetch}`;
-    const matchRows = db.prepare(matchSql).all(terms);
+    let matchRows;
+    try {
+        matchRows = db.prepare(matchSql).all(terms);
+    } catch (err) {
+        throw searchError(err, terms, scope);
+    }
     const hits = new Map(matchRows.map((r)=>[
             r.path,
             r.hit
@@ -32,20 +68,26 @@ export function find(db, cfg, terms, opts = {}) {
             via: 'match'
         });
     });
-    if (featureEnabled(cfg, 'links') && matchRows.length > 0) {
+    const edges = featureEnabled(cfg, 'links') && matchRows.length > 0 ? linkEdges(db) : [];
+    if (edges.length > 0) {
+        // `linked` gates the label only, not the score: PPR restart mass gives every seed a
+        // nonzero rank even without an incident edge, which is not link evidence — but dropping
+        // that mass from the score list reweights fusion toward connectivity and measurably
+        // wrecks ranking on link-dense corpora (FEVER hit@10 0.997 -> 0.907; fusion-tuning.md).
+        const linked = new Set(edges.flat());
         const nodes = db.prepare('SELECT "path" FROM frontmatter').all().map((r)=>r.path);
         const seeds = new Map(matchRows.map((r, i)=>[
                 r.path,
                 1 / (i + 1)
             ]));
         const ranked = [
-            ...personalizedRank(nodes, linkEdges(db), seeds)
+            ...personalizedRank(nodes, edges, seeds)
         ].filter(([, score])=>score > 1e-9).sort((a, b)=>b[1] - a[1]).slice(0, fetch);
         ranked.forEach(([path], i)=>{
             const existing = candidates.get(path);
             if (existing) {
                 existing.score += 1 / (RRF_K + i);
-                existing.via = 'match+link';
+                if (linked.has(path)) existing.via = 'match+link';
             } else {
                 candidates.set(path, {
                     score: 1 / (RRF_K + i),
@@ -54,12 +96,34 @@ export function find(db, cfg, terms, opts = {}) {
             }
         });
     }
-    db.exec('CREATE TEMP TABLE IF NOT EXISTS _find ("path" TEXT PRIMARY KEY, score REAL, via TEXT, hit TEXT)');
-    db.exec('DELETE FROM _find');
-    const insert = db.prepare('INSERT INTO _find ("path", score, via, hit) VALUES (?, ?, ?, ?)');
-    for (const [path, c] of candidates)insert.run(path, c.score, c.via, (_hits_get = hits.get(path)) !== null && _hits_get !== void 0 ? _hits_get : null);
-    const where = opts.where ? `WHERE ${opts.where}` : '';
-    return db.prepare(`SELECT f."path" AS path, content.title, content.summary, _find.hit, _find.via, round(_find.score, 4) AS score
+    // Vector expansion, invoked only: a third RRF list at the swept flat-region constants
+    // (weight 1, pool = fetch). Each row carries its best chunk's line range.
+    const chunkLines = new Map();
+    const chunkSimilarity = new Map();
+    if (opts.semantic) {
+        const vec = await semanticCandidates(db, cfg, terms, fetch);
+        vec.forEach(({ path, lines, similarity }, i)=>{
+            chunkLines.set(path, lines);
+            chunkSimilarity.set(path, similarity);
+            const existing = candidates.get(path);
+            if (existing) {
+                existing.score += 1 / (RRF_K + i);
+                existing.via = `${existing.via}+vector`;
+            } else {
+                candidates.set(path, {
+                    score: 1 / (RRF_K + i),
+                    via: 'vector'
+                });
+            }
+        });
+    }
+    db.exec('DROP TABLE IF EXISTS _find');
+    db.exec('CREATE TEMP TABLE _find ("path" TEXT PRIMARY KEY, score REAL, via TEXT, hit TEXT, lines TEXT, similarity REAL)');
+    const insert = db.prepare('INSERT INTO _find ("path", score, via, hit, lines, similarity) VALUES (?, ?, ?, ?, ?, ?)');
+    for (const [path, c] of candidates)insert.run(path, c.score, c.via, (_hits_get = hits.get(path)) !== null && _hits_get !== void 0 ? _hits_get : null, (_chunkLines_get = chunkLines.get(path)) !== null && _chunkLines_get !== void 0 ? _chunkLines_get : null, (_chunkSimilarity_get = chunkSimilarity.get(path)) !== null && _chunkSimilarity_get !== void 0 ? _chunkSimilarity_get : null);
+    const where = scope ? `WHERE ${scope}` : '';
+    const linesCol = opts.semantic ? ', _find.lines, _find.similarity' : '';
+    return db.prepare(`SELECT f."path" AS path, content.title, content.summary, _find.hit, _find.via, round(_find.score, 4) AS score${linesCol}
        FROM _find JOIN frontmatter f ON f."path" = _find."path" JOIN content ON content.path = _find."path"
        ${where} ORDER BY _find.score DESC LIMIT ?`).all(k);
 }
@@ -74,10 +138,15 @@ export function mapTree(db, cfg) {
     const docs = db.prepare('SELECT COUNT(*) AS count, COALESCE(SUM("_size"), 0) AS bytes FROM frontmatter').get();
     const columns = db.prepare('PRAGMA table_info(frontmatter)').all().map((r)=>r.name).filter((name)=>!INTERNAL_COLUMNS.has(name));
     const allFields = columns.map((name)=>{
-        const { n } = db.prepare(`SELECT COUNT("${name.split('"').join('""')}") AS n FROM frontmatter`).get();
+        const quoted = `"${name.split('"').join('""')}"`;
+        // Observed storage class, not a declared one: these columns are added dynamically and
+        // SQLite types per value, so a field can be text in most notes and numeric in a few.
+        // Listing every distinct type makes both the type and any drift visible.
+        const { n, types } = db.prepare(`SELECT COUNT(${quoted}) AS n, GROUP_CONCAT(DISTINCT typeof(${quoted})) AS types FROM frontmatter WHERE ${quoted} IS NOT NULL`).get();
         return {
             field: name,
-            coverage: n
+            coverage: n,
+            type: types !== null && types !== void 0 ? types : ''
         };
     }).sort((a, b)=>b.coverage - a.coverage);
     const fields = allFields.slice(0, 20);
@@ -87,6 +156,7 @@ export function mapTree(db, cfg) {
         docs,
         fields,
         fieldsTotal: allFields.length,
+        features: featureStates(cfg),
         hubs,
         recent
     };
@@ -134,6 +204,10 @@ export function peek(db, cfg, pathArg) {
         unresolved: unresolved.slice(0, PEEK_LINK_LIMIT),
         outboundTotal: outbound.length,
         backlinksTotal,
-        unresolvedTotal: unresolved.length
+        unresolvedTotal: unresolved.length,
+        off: [
+            'sections',
+            'links'
+        ].filter((name)=>!featureEnabled(cfg, name))
     };
 }