sensemaking 0.7.2 → 0.9.0

package/README.md

@@ -1,7 +1,8 @@
 # sensemaking
 
-SQL over a tree of markdown notes: filter by frontmatter, search the prose, follow the links,
-read the structure. A CLI that starts, answers, and exits — no server, no build step.
+Query and search a tree of markdown notes: SQL over frontmatter and links, ranked search over
+the prose — words, links, and meaning fused. A CLI that starts, answers, and exits — no
+server, no build step.
 
 ## Problem
 
@@ -22,14 +23,14 @@ cd your-notes && sense init
 
 ```bash
 sense map                                        # orient: fields, hub notes, recent changes
-sense find "revenue OR earnings" --k 10          # locate: ranked references with excerpts
+sense search "revenue OR earnings" --k 10        # locate: words + links + meaning, one ranked list
 sense peek notes/q3-report.md                    # structure: outline + links, before reading
 sense query "SELECT path FROM frontmatter WHERE has(tags, ?)" urgent
 ```
 
 ## Model
 
-Every file becomes rows in four tables:
+Every file becomes rows in these tables, plus whatever an enabled feature adds of its own:
 
 | table | holds | for |
 |---|---|---|
@@ -43,8 +44,9 @@ afterward through the filesystem, scoped to the line ranges `peek` returns. This
 [just-in-time context pattern](https://www.anthropic.com/engineering/effective-context-engineering-for-ai-agents):
 the agent holds lightweight identifiers and loads payloads only when needed.
 
-Measured costs: a `find` row is ~20–40 tokens, a `peek` is ~17% of reading the file, and the
-section it points at is a direct `Read` range.
+Output size is a contract, measured per release ([BENCHMARKING.md](BENCHMARKING.md)): `map` is
+fixed-size, a `find` row is tens of tokens, and a `peek` stays flat however large the note is —
+so what it saves over reading grows with the file, and a small note is cheaper read whole.
 
 ```sql
 -- filter and search compose in one query
@@ -54,74 +56,93 @@ WHERE f.status = 'active' AND content MATCH 'revenue'
 ORDER BY bm25(content, 10.0, 5.0, 1.0) LIMIT 10
 ```
 
-## Verbs
+## Commands
 
-| verb | does |
+| command | does |
 |---|---|
 | `map` | doc count, frontmatter field coverage, top hubs by link rank, recent changes |
-| `find "<terms>" [--where "<sql>"] [--k n] [--semantic]` | BM25 + link-graph expansion, fused; `via` marks each row `match`, `link`, or `match+link` |
-| `peek <path>` | frontmatter + heading outline (`[L143-162, ~380t]`) + links both ways, capped at 20 per list |
-| `query "<sql>" [params...]` | ad-hoc SQL over all four tables; `?` binds positional args |
-
-`find` seeds a personalized-PageRank walk with the BM25 matches, so a note that never contains
-the terms but is linked from ones that do still surfaces. Terms pass verbatim to FTS5 `MATCH`
-(bare words AND-join; operators are yours to write). `--where` takes a frontmatter condition
-against alias `f`. On trees with the `embed` feature, `--semantic` additionally expands by
-meaning — explicit per query, never silent; rows it adds are labeled `via: vector` and carry a
-`lines` column pointing at the best-matching section.
-`--format json` on any verb returns structured output.
+| `search "<text>" [--preset name] [--include glob] [--where "<sql>"] [--k n] [--lexical]` | words + links + vectors, one fused ranked list; `via` labels each row's evidence |
+| `peek <path>` | frontmatter + heading outline (`[L143-162, ~380t]`) + links both ways — first 20 per list, each with its total; the `links` table has the rest |
+| `query "<sql>" [params...]` | ad-hoc SQL over all the tables; `?` binds positional args |
+| `<name> [params...]` | run a query saved in the config; `--list` names them |
+| `init` | write a starter `sense.config.json` |
+| `status` | index location, doc count, per-preset coverage, watcher heartbeat |
+| `check` | run every saved query and search, so a broken one fails here instead of mid-task |
+| `rebuild` | delete the cache and re-crawl |
+| `watch` | keep the index warm in the background (optional; see [WATCH.md](WATCH.md)) |
+
+`search` runs one text through every engine its scope has — FTS5 word match (BM25-ranked,
+bare words AND-join, operators are yours), a personalized-PageRank walk over the link graph,
+and vector similarity — fused into one list. `via` labels each row's evidence (`match`,
+`link`, `vector`, combinations); `similarity` is the cosine against the best-matching chunk;
+`lines` points at the section that earned the row — a direct read range. Rows that only
+vectors produced are the "these words aren't in the tree, this is what's near in meaning"
+signal. `--preset` picks a named settings bundle from the config, `--lexical` skips vectors
+for one command, `--where` filters on frontmatter. `--format json` on any command returns
+structured output; `--version` and `--help` do what they say.
 
 ## Config
 
 `sense init` writes `sense.config.json`; discovery walks up from cwd like git
-(`--config <path>` overrides). Name reusable queries and run them as `sense <name> [params...]`:
+(`--config <path>` overrides). Three keys: **presets** (named, self-contained setting
+bundles), **queries** (saved commands), and the version.
 
 ```json
 {
   "$schema": "https://unpkg.com/sensemaking/schema.json",
-  "version": 2,
-  "scan": { "include": ["**/*.md"] },
-  "features": { "links": true, "sections": true, "rank": true },
+  "version": 3,
+  "presets": {
+    "default": { "include": ["**/*.md"], "k": 10 },
+    "raw":     { "include": ["raw/**/*.md"], "k": 5, "semantic": false }
+  },
   "queries": {
-    "by-tag": "SELECT path, title FROM frontmatter WHERE has(tags, ?) ORDER BY path"
+    "dead-links": "SELECT src, target FROM links WHERE dst IS NULL",
+    "by-tag": "SELECT path, title FROM frontmatter WHERE has(tags, ?) ORDER BY path",
+    "hot": { "search": "pricing OR billing", "preset": "raw" }
   }
 }
 ```
 
-Features toggle independently; disabling one degrades its verb instead of failing it (`find`
-goes BM25-only without `links`, `map` drops hubs without `rank`, `peek` drops the outline
-without `sections`). Toggling rebuilds the cache. Older config versions auto-migrate on load,
-noted on stderr.
-
-`embed` is the one opt-in feature (absent = off; most trees don't need vectors): `"embed": true`
-uses the built-in static model (downloaded to `~/.cache/sensemaking` on first use, never in the
-package), or `{ "model", "type": "static"|"api", "url", "key" }` points at any Model2Vec model,
-local path, or OpenAI-compatible endpoint (Ollama, LM Studio, hosted). With it on, vectors stay
-fresh at reconcile and `find --semantic` uses them; default `find` results are unchanged.
-
-## Reference
-
-- `has(field, value)` — the one custom SQL function: array membership on JSON-array fields,
-  substring on strings (so `has(f.status, 'active')` also matches `inactive`), false on missing
-  keys. Exact matches: `=` for scalars, `EXISTS (SELECT 1 FROM json_each(f.tags) WHERE value = ?)`
-  for array members.
-- Reserved frontmatter keys (dropped with a warning): `path`, `_mtime`, `_size`, `_rank`,
-  `content`, `links`, `sections`.
-- FTS5 syntax in `MATCH`: `a OR b`, `"phrase"`, `pref*`, `NEAR(a b, 5)`, `summary: term`.
-  Stemmed; markdown is stripped at index time. `bm25(content, 10.0, 5.0, 1.0)` weights title
-  over summary over body.
-- A one-line `summary:` in frontmatter is both a selectable column and a weighted search field.
-- Exit codes: `0` ok, `1` error (SQLite message verbatim), `2` usage.
-- Frontmatter parsing is lenient: syntax errors (an alias starting with `@`, say) are per-file
-  warnings and the values are still indexed.
+A preset bundles `include`/`exclude` globs (which files), `k` (result count), `semantic`
+(vectors, on unless `false`), and `where` (a standing SQL filter). Bare commands use
+`default`; `--preset` names another; flags override single fields. **Indexing derives from
+the presets**: a file is indexed if any preset includes it, and embedded if any covering
+preset has semantic on — so a `semantic: false` preset's files cost no vectors, and
+`status` shows each preset's coverage. Editing a preset rebuilds the cache and says which
+preset caused it.
+
+A query entry is a SQL string (`?` binds positional args) or a saved search with its
+settings baked in — `sense hot` needs no flags. `sense check` runs them all, so a typo'd
+column fails at check time instead of mid-task. Older config versions auto-migrate on
+load, noted on stderr.
+
+Vectors use the built-in static model (downloaded to `~/.cache/sensemaking` on first use,
+never in the package); an optional top-level `"embed": { "model", "type", "url", "key" }`
+block points at any Model2Vec model, local path, or OpenAI-compatible endpoint (Ollama,
+LM Studio, hosted). Embedding happens on the first semantic search, with progress.
+
+`has(field, value)` is the one custom SQL function: array membership on JSON-array fields,
+substring on strings, false on missing keys. Frontmatter parsing is lenient — syntax errors are
+per-file warnings and the values are still indexed, so one bad note never costs you the crawl.
 
 ## Scale
 
 Every query starts with a freshness check against the cache in `.sense/`; only changed files are
-re-parsed. Measured per release in [BENCHMARKING.md](BENCHMARKING.md) — at 6,566 notes: full
-crawl ~5 s, warm query ~100 ms, one-file change ~200 ms. `sense watch` (optional) moves
-re-parsing into the background — see [WATCH.md](WATCH.md). `sense rebuild` deletes the cache and
-re-crawls.
+re-parsed. What to expect as a tree grows:
+
+- **Work is linear in note count** — crawl, reconcile, and the freshness check every invocation
+  pays. That check is the floor cost of a query and the first thing to watch on a large tree.
+- **Output is flat.** `map`, `peek`, and a `find` row cost the same on a small tree as a large
+  one: context cost is bounded by what you ask for, not by how much there is.
+- **Bulk changes are paid by whoever queries next.** `sense watch` moves that re-parse into the
+  background ([WATCH.md](WATCH.md)) — it changes latency, never answers, since every query
+  reconciles for itself. `sense rebuild` starts the cache over.
+
+These are release gates rather than hopes: every release regenerates the numbers on pinned
+corpora spanning a 4x range in note count plus a stress tree that packs the worst measured
+shapes into one place — a megabyte-scale note, heading-dense outlines, dense link graphs,
+hundreds of frontmatter fields. A row that grows faster than linearly, or a token count
+that grows at all, blocks the release. Current figures: [BENCHMARKING.md](BENCHMARKING.md).
 
 ## For AI agents
 
@@ -129,18 +150,23 @@ re-crawls.
 npx skills add kmalakoff/sensemaking   # -g for global, -a claude-code to target
 ```
 
-The skill teaches the descent: `map` to orient, `find` to locate, `peek` before reading,
-`Read` line ranges instead of files.
+Two skills: `sense` for querying a tree — what each command is for, FTS5 syntax, reading the
+`via`/`score`/`similarity` columns, worked examples — and `sense-setup` for making one, where
+features, frontmatter conventions, and note size are decisions with consequences either way.
 
 ## Prior art
 
 - [Effective context engineering for AI agents](https://www.anthropic.com/engineering/effective-context-engineering-for-ai-agents)
   (Anthropic): agents should hold lightweight identifiers — file paths, links — and load payloads
-  just in time, because context is a finite resource. The four verbs implement that pattern as a CLI.
+  just in time, because context is a finite resource. The commands implement that pattern as a CLI.
 - [llm-wiki](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f) (Karpathy): an
   agent-maintained wiki navigated by an `index.md` and links, which he notes needs real search
   infrastructure past a few hundred pages. `sense map` derives that index from the notes instead of
   maintaining it; `find` is the hybrid local search it calls for.
+- Agent memory patterns — llm-wiki's raw/wiki split, Claude Code's dreaming-style nightly
+  consolidation — are trees of small notes with metadata, links, and layers of differing
+  authority. sense is the query layer such patterns need (filter by metadata and age, scope
+  by layer, surface near-duplicates semantically), not an implementation of any one of them.
 
 ## Alternatives
 
@@ -148,13 +174,16 @@ The skill teaches the descent: `map` to orient, `find` to locate, `peek` before
   query it headless.
 - **Index-on-build tools (MarkdownDB)** — query a snapshot; `sense` reconciles on every query.
 - **Note CLIs (zk)** — fixed schema; `sense` filters on arbitrary frontmatter.
-- **RAG / vector stores** — similarity can't express `WHERE status = 'active'`. Here vectors are
-  the optional `embed` feature: same SQLite file, filters compose, expansion is explicit per
-  query (`find --semantic`) and labeled — no second store, no daemon, no native builds.
+- **RAG / vector stores** — similarity can't express `WHERE status = 'active'`. Here vectors
+  are one signal inside `search`: same SQLite file, filters compose, every row labels its
+  evidence (`via`), and a preset turns vectors off per layer of the tree — no second store,
+  no daemon, no native builds.
 
 Dependencies: [yaml](https://github.com/eemeli/yaml),
 [remove-markdown](https://github.com/zuchka/remove-markdown),
-[fast-glob](https://github.com/mrmlnc/fast-glob), and Node's built-in SQLite. No native builds.
+[fast-glob](https://github.com/mrmlnc/fast-glob),
+[@huggingface/tokenizers](https://github.com/huggingface/tokenizers.js) (pure JS), and Node's
+built-in SQLite. No native builds.
 
 ## License
 

package/dist/cjs/cli/check.js

@@ -0,0 +1,367 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", {
+    value: true
+});
+Object.defineProperty(exports, "default", {
+    enumerable: true,
+    get: function() {
+        return _default;
+    }
+});
+var _commandsts = require("../commands.js");
+var _dbts = require("../db.js");
+var _outputts = require("../output.js");
+var _sharedts = require("./shared.js");
+function _array_like_to_array(arr, len) {
+    if (len == null || len > arr.length) len = arr.length;
+    for(var i = 0, arr2 = new Array(len); i < len; i++)arr2[i] = arr[i];
+    return arr2;
+}
+function _array_with_holes(arr) {
+    if (Array.isArray(arr)) return arr;
+}
+function asyncGeneratorStep(gen, resolve, reject, _next, _throw, key, arg) {
+    try {
+        var info = gen[key](arg);
+        var value = info.value;
+    } catch (error) {
+        reject(error);
+        return;
+    }
+    if (info.done) {
+        resolve(value);
+    } else {
+        Promise.resolve(value).then(_next, _throw);
+    }
+}
+function _async_to_generator(fn) {
+    return function() {
+        var self = this, args = arguments;
+        return new Promise(function(resolve, reject) {
+            var gen = fn.apply(self, args);
+            function _next(value) {
+                asyncGeneratorStep(gen, resolve, reject, _next, _throw, "next", value);
+            }
+            function _throw(err) {
+                asyncGeneratorStep(gen, resolve, reject, _next, _throw, "throw", err);
+            }
+            _next(undefined);
+        });
+    };
+}
+function _iterable_to_array_limit(arr, i) {
+    var _i = arr == null ? null : typeof Symbol !== "undefined" && arr[Symbol.iterator] || arr["@@iterator"];
+    if (_i == null) return;
+    var _arr = [];
+    var _n = true;
+    var _d = false;
+    var _s, _e;
+    try {
+        for(_i = _i.call(arr); !(_n = (_s = _i.next()).done); _n = true){
+            _arr.push(_s.value);
+            if (i && _arr.length === i) break;
+        }
+    } catch (err) {
+        _d = true;
+        _e = err;
+    } finally{
+        try {
+            if (!_n && _i["return"] != null) _i["return"]();
+        } finally{
+            if (_d) throw _e;
+        }
+    }
+    return _arr;
+}
+function _non_iterable_rest() {
+    throw new TypeError("Invalid attempt to destructure non-iterable instance.\\nIn order to be iterable, non-array objects must have a [Symbol.iterator]() method.");
+}
+function _sliced_to_array(arr, i) {
+    return _array_with_holes(arr) || _iterable_to_array_limit(arr, i) || _unsupported_iterable_to_array(arr, i) || _non_iterable_rest();
+}
+function _type_of(obj) {
+    "@swc/helpers - typeof";
+    return obj && typeof Symbol !== "undefined" && obj.constructor === Symbol ? "symbol" : typeof obj;
+}
+function _unsupported_iterable_to_array(o, minLen) {
+    if (!o) return;
+    if (typeof o === "string") return _array_like_to_array(o, minLen);
+    var n = Object.prototype.toString.call(o).slice(8, -1);
+    if (n === "Object" && o.constructor) n = o.constructor.name;
+    if (n === "Map" || n === "Set") return Array.from(n);
+    if (n === "Arguments" || /^(?:Ui|I)nt(?:8|16|32)(?:Clamped)?Array$/.test(n)) return _array_like_to_array(o, minLen);
+}
+function _ts_generator(thisArg, body) {
+    var f, y, t, _ = {
+        label: 0,
+        sent: function() {
+            if (t[0] & 1) throw t[1];
+            return t[1];
+        },
+        trys: [],
+        ops: []
+    }, g = Object.create((typeof Iterator === "function" ? Iterator : Object).prototype), d = Object.defineProperty;
+    return d(g, "next", {
+        value: verb(0)
+    }), d(g, "throw", {
+        value: verb(1)
+    }), d(g, "return", {
+        value: verb(2)
+    }), typeof Symbol === "function" && d(g, Symbol.iterator, {
+        value: function() {
+            return this;
+        }
+    }), g;
+    function verb(n) {
+        return function(v) {
+            return step([
+                n,
+                v
+            ]);
+        };
+    }
+    function step(op) {
+        if (f) throw new TypeError("Generator is already executing.");
+        while(g && (g = 0, op[0] && (_ = 0)), _)try {
+            if (f = 1, y && (t = op[0] & 2 ? y["return"] : op[0] ? y["throw"] || ((t = y["return"]) && t.call(y), 0) : y.next) && !(t = t.call(y, op[1])).done) return t;
+            if (y = 0, t) op = [
+                op[0] & 2,
+                t.value
+            ];
+            switch(op[0]){
+                case 0:
+                case 1:
+                    t = op;
+                    break;
+                case 4:
+                    _.label++;
+                    return {
+                        value: op[1],
+                        done: false
+                    };
+                case 5:
+                    _.label++;
+                    y = op[1];
+                    op = [
+                        0
+                    ];
+                    continue;
+                case 7:
+                    op = _.ops.pop();
+                    _.trys.pop();
+                    continue;
+                default:
+                    if (!(t = _.trys, t = t.length > 0 && t[t.length - 1]) && (op[0] === 6 || op[0] === 2)) {
+                        _ = 0;
+                        continue;
+                    }
+                    if (op[0] === 3 && (!t || op[1] > t[0] && op[1] < t[3])) {
+                        _.label = op[1];
+                        break;
+                    }
+                    if (op[0] === 6 && _.label < t[1]) {
+                        _.label = t[1];
+                        t = op;
+                        break;
+                    }
+                    if (t && _.label < t[2]) {
+                        _.label = t[2];
+                        _.ops.push(op);
+                        break;
+                    }
+                    if (t[2]) _.ops.pop();
+                    _.trys.pop();
+                    continue;
+            }
+            op = body.call(thisArg, _);
+        } catch (e) {
+            op = [
+                6,
+                e
+            ];
+            y = 0;
+        } finally{
+            f = t = 0;
+        }
+        if (op[0] & 5) throw op[1];
+        return {
+            value: op[0] ? op[1] : void 0,
+            done: true
+        };
+    }
+}
+// A saved query that returns zero rows looks identical to a true empty result, so a broken
+// one can sit unnoticed -- one field report had `WHERE flag = 'true'` against a numeric
+// column reading as "nothing tagged yet" for hours. Preparing each query catches syntax and
+// unknown-column errors without executing it; queries that take no parameters are also run
+// for a row count. Parameterised queries are validated but not counted: inventing arguments
+// would report a count for a query nobody ran. Saved searches run lexically with k=1 -- that
+// validates the `where` fragment, the FTS5 terms, and the `preset` name (resolveSearch throws
+// on an unknown one, the same breakage class prepare() catches for SQL strings) -- but never
+// semantically (a semantic pass costs model time and, on api trees, network). Whether a
+// result being empty is good or bad is the reader's judgment -- there is no assertion path.
+var check = function check(ctx) {
+    return _async_to_generator(function() {
+        var _cfg_queries, cfg, _open, db, warnings, rows, failed, _iteratorNormalCompletion, _didIteratorError, _iteratorError, _iterator, _step, _step_value, name, entry, _sql_match, count, err, sql, params, statement, count1, err1;
+        return _ts_generator(this, function(_state) {
+            switch(_state.label){
+                case 0:
+                    cfg = ctx.resolveConfig();
+                    _open = (0, _dbts.open)(cfg), db = _open.db, warnings = _open.warnings;
+                    (0, _sharedts.printWarnings)(warnings);
+                    rows = [];
+                    failed = 0;
+                    _iteratorNormalCompletion = true, _didIteratorError = false, _iteratorError = undefined;
+                    _state.label = 1;
+                case 1:
+                    _state.trys.push([
+                        1,
+                        10,
+                        11,
+                        12
+                    ]);
+                    _iterator = Object.entries((_cfg_queries = cfg.queries) !== null && _cfg_queries !== void 0 ? _cfg_queries : {})[Symbol.iterator]();
+                    _state.label = 2;
+                case 2:
+                    if (!!(_iteratorNormalCompletion = (_step = _iterator.next()).done)) return [
+                        3,
+                        9
+                    ];
+                    _step_value = _sliced_to_array(_step.value, 2), name = _step_value[0], entry = _step_value[1];
+                    if (!((typeof entry === "undefined" ? "undefined" : _type_of(entry)) === 'object' && entry !== null && 'search' in entry)) return [
+                        3,
+                        7
+                    ];
+                    _state.label = 3;
+                case 3:
+                    _state.trys.push([
+                        3,
+                        5,
+                        ,
+                        6
+                    ]);
+                    return [
+                        4,
+                        (0, _commandsts.search)(db, cfg, entry.search, {
+                            k: 1,
+                            where: entry.where,
+                            preset: entry.preset,
+                            include: entry.include,
+                            semantic: false
+                        })
+                    ];
+                case 4:
+                    count = _state.sent().length;
+                    rows.push({
+                        query: name,
+                        params: '—',
+                        rows: '—',
+                        status: count === 0 ? 'ok, but matches 0 notes (lexical probe)' : 'ok (probe run with k=1)'
+                    });
+                    return [
+                        3,
+                        6
+                    ];
+                case 5:
+                    err = _state.sent();
+                    failed++;
+                    rows.push({
+                        query: name,
+                        params: '—',
+                        rows: '—',
+                        status: "FAILED: ".concat(err.message)
+                    });
+                    return [
+                        3,
+                        6
+                    ];
+                case 6:
+                    return [
+                        3,
+                        8
+                    ];
+                case 7:
+                    sql = typeof entry === 'string' ? entry : entry.sql;
+                    params = ((_sql_match = sql.match(/\?/g)) !== null && _sql_match !== void 0 ? _sql_match : []).length;
+                    try {
+                        statement = db.prepare(sql);
+                        if (params > 0) {
+                            rows.push({
+                                query: name,
+                                params: params,
+                                rows: '—',
+                                status: 'ok (not run: needs parameters)'
+                            });
+                            return [
+                                3,
+                                8
+                            ];
+                        }
+                        count1 = statement.all().length;
+                        rows.push({
+                            query: name,
+                            params: params,
+                            rows: count1,
+                            status: count1 === 0 ? 'ok, but returns 0 rows' : 'ok'
+                        });
+                    } catch (err) {
+                        failed++;
+                        rows.push({
+                            query: name,
+                            params: params,
+                            rows: '—',
+                            status: "FAILED: ".concat(err.message)
+                        });
+                    }
+                    _state.label = 8;
+                case 8:
+                    _iteratorNormalCompletion = true;
+                    return [
+                        3,
+                        2
+                    ];
+                case 9:
+                    return [
+                        3,
+                        12
+                    ];
+                case 10:
+                    err1 = _state.sent();
+                    _didIteratorError = true;
+                    _iteratorError = err1;
+                    return [
+                        3,
+                        12
+                    ];
+                case 11:
+                    try {
+                        if (!_iteratorNormalCompletion && _iterator.return != null) {
+                            _iterator.return();
+                        }
+                    } finally{
+                        if (_didIteratorError) {
+                            throw _iteratorError;
+                        }
+                    }
+                    return [
+                        7
+                    ];
+                case 12:
+                    db.close();
+                    if (rows.length === 0) {
+                        console.log('no saved queries in config');
+                        return [
+                            2
+                        ];
+                    }
+                    (0, _outputts.printRows)(rows, ctx.format);
+                    if (failed > 0) process.exit(1);
+                    return [
+                        2
+                    ];
+            }
+        });
+    })();
+};
+var _default = check;
+/* CJS INTEROP */ if (exports.__esModule && exports.default) { try { Object.defineProperty(exports.default, '__esModule', { value: true }); for (var key in exports) { exports.default[key] = exports[key]; } } catch (_) {}; module.exports = exports.default; }

package/dist/cjs/cli/check.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/cli/check.ts"],"sourcesContent":["import { search } from '../commands.ts';\nimport { open } from '../db.ts';\nimport type { Row } from '../output.ts';\nimport { printRows } from '../output.ts';\nimport { printWarnings } from './shared.ts';\nimport type { Command } from './types.ts';\n\n// A saved query that returns zero rows looks identical to a true empty result, so a broken\n// one can sit unnoticed -- one field report had `WHERE flag = 'true'` against a numeric\n// column reading as \"nothing tagged yet\" for hours. Preparing each query catches syntax and\n// unknown-column errors without executing it; queries that take no parameters are also run\n// for a row count. Parameterised queries are validated but not counted: inventing arguments\n// would report a count for a query nobody ran. Saved searches run lexically with k=1 -- that\n// validates the `where` fragment, the FTS5 terms, and the `preset` name (resolveSearch throws\n// on an unknown one, the same breakage class prepare() catches for SQL strings) -- but never\n// semantically (a semantic pass costs model time and, on api trees, network). Whether a\n// result being empty is good or bad is the reader's judgment -- there is no assertion path.\nconst check: Command = async (ctx) => {\n  const cfg = ctx.resolveConfig();\n  const { db, warnings } = open(cfg);\n  printWarnings(warnings);\n\n  const rows: Row[] = [];\n  let failed = 0;\n  for (const [name, entry] of Object.entries(cfg.queries ?? {})) {\n    if (typeof entry === 'object' && entry !== null && 'search' in entry) {\n      try {\n        const count = (await search(db, cfg, entry.search, { k: 1, where: entry.where, preset: entry.preset, include: entry.include, semantic: false })).length;\n        rows.push({ query: name, params: '—', rows: '—', status: count === 0 ? 'ok, but matches 0 notes (lexical probe)' : 'ok (probe run with k=1)' });\n      } catch (err) {\n        failed++;\n        rows.push({ query: name, params: '—', rows: '—', status: `FAILED: ${(err as Error).message}` });\n      }\n      continue;\n    }\n    const sql = typeof entry === 'string' ? entry : entry.sql;\n    const params = (sql.match(/\\?/g) ?? []).length;\n    try {\n      const statement = db.prepare(sql);\n      if (params > 0) {\n        rows.push({ query: name, params, rows: '—', status: 'ok (not run: needs parameters)' });\n        continue;\n      }\n      const count = (statement.all() as unknown[]).length;\n      rows.push({ query: name, params, rows: count, status: count === 0 ? 'ok, but returns 0 rows' : 'ok' });\n    } catch (err) {\n      failed++;\n      rows.push({ query: name, params, rows: '—', status: `FAILED: ${(err as Error).message}` });\n    }\n  }\n\n  db.close();\n  if (rows.length === 0) {\n    console.log('no saved queries in config');\n    return;\n  }\n  printRows(rows, ctx.format);\n  if (failed > 0) process.exit(1);\n};\nexport default check;\n"],"names":["check","ctx","cfg","open","db","warnings","rows","failed","name","entry","sql","count","err","params","statement","resolveConfig","printWarnings","Object","entries","queries","search","k","where","preset","include","semantic","length","push","query","status","message","match","prepare","all","close","console","log","printRows","format","process","exit"],"mappings":";;;;+BA2DA;;;eAAA;;;0BA3DuB;oBACF;wBAEK;wBACI;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAG9B,2FAA2F;AAC3F,wFAAwF;AACxF,4FAA4F;AAC5F,2FAA2F;AAC3F,4FAA4F;AAC5F,6FAA6F;AAC7F,8FAA8F;AAC9F,6FAA6F;AAC7F,wFAAwF;AACxF,4FAA4F;AAC5F,IAAMA,QAAiB,eAAOC;;YAOeC,cANrCA,KACmBC,OAAjBC,IAAIC,UAGNC,MACFC,QACC,2BAAA,mBAAA,gBAAA,WAAA,oBAAOC,MAAMC,OAYAC,YATNC,OAECC,KAMLF,KACAG,QAEEC,WAKAH;;;;oBAzBJT,MAAMD,IAAIc,aAAa;oBACJZ,QAAAA,IAAAA,UAAI,EAACD,MAAtBE,KAAiBD,MAAjBC,IAAIC,WAAaF,MAAbE;oBACZW,IAAAA,uBAAa,EAACX;oBAERC;oBACFC,SAAS;oBACR,kCAAA,2BAAA;;;;;;;;;oBAAA,YAAuBU,OAAOC,OAAO,EAAChB,eAAAA,IAAIiB,OAAO,cAAXjB,0BAAAA,eAAe,CAAC;;;2BAAtD,6BAAA,QAAA;;;;mDAAA,iBAAOM,uBAAMC;yBACZ,CAAA,CAAA,OAAOA,sCAAP,SAAOA,MAAI,MAAM,YAAYA,UAAU,QAAQ,YAAYA,KAAI,GAA/D;;;;;;;;;;;;oBAEe;;wBAAMW,IAAAA,kBAAM,EAAChB,IAAIF,KAAKO,MAAMW,MAAM,EAAE;4BAAEC,GAAG;4BAAGC,OAAOb,MAAMa,KAAK;4BAAEC,QAAQd,MAAMc,MAAM;4BAAEC,SAASf,MAAMe,OAAO;4BAAEC,UAAU;wBAAM;;;oBAAvId,QAAQ,AAAC,cAAkIe,MAAM;oBACvJpB,KAAKqB,IAAI,CAAC;wBAAEC,OAAOpB;wBAAMK,QAAQ;wBAAKP,MAAM;wBAAKuB,QAAQlB,UAAU,IAAI,4CAA4C;oBAA0B;;;;;;oBACtIC;oBACPL;oBACAD,KAAKqB,IAAI,CAAC;wBAAEC,OAAOpB;wBAAMK,QAAQ;wBAAKP,MAAM;wBAAKuB,QAAQ,AAAC,WAAiC,OAAvB,AAACjB,IAAckB,OAAO;oBAAG;;;;;;oBAE/F;;;;;oBAEIpB,MAAM,OAAOD,UAAU,WAAWA,QAAQA,MAAMC,GAAG;oBACnDG,SAAS,EAACH,aAAAA,IAAIqB,KAAK,CAAC,oBAAVrB,wBAAAA,iBAAwBgB,MAAM;oBAC9C,IAAI;wBACIZ,YAAYV,GAAG4B,OAAO,CAACtB;wBAC7B,IAAIG,SAAS,GAAG;4BACdP,KAAKqB,IAAI,CAAC;gCAAEC,OAAOpB;gCAAMK,QAAAA;gCAAQP,MAAM;gCAAKuB,QAAQ;4BAAiC;4BACrF;;;;wBACF;wBACMlB,SAAQ,AAACG,UAAUmB,GAAG,GAAiBP,MAAM;wBACnDpB,KAAKqB,IAAI,CAAC;4BAAEC,OAAOpB;4BAAMK,QAAAA;4BAAQP,MAAMK;4BAAOkB,QAAQlB,WAAU,IAAI,2BAA2B;wBAAK;oBACtG,EAAE,OAAOC,KAAK;wBACZL;wBACAD,KAAKqB,IAAI,CAAC;4BAAEC,OAAOpB;4BAAMK,QAAAA;4BAAQP,MAAM;4BAAKuB,QAAQ,AAAC,WAAiC,OAAvB,AAACjB,IAAckB,OAAO;wBAAG;oBAC1F;;;oBAxBG;;;;;;;;;;;;oBAAA;oBAAA;;;;;;;6BAAA,6BAAA;4BAAA;;;4BAAA;kCAAA;;;;;;;oBA2BL1B,GAAG8B,KAAK;oBACR,IAAI5B,KAAKoB,MAAM,KAAK,GAAG;wBACrBS,QAAQC,GAAG,CAAC;wBACZ;;;oBACF;oBACAC,IAAAA,mBAAS,EAAC/B,MAAML,IAAIqC,MAAM;oBAC1B,IAAI/B,SAAS,GAAGgC,QAAQC,IAAI,CAAC;;;;;;IAC/B;;IACA,WAAexC"}

package/dist/cjs/{commands → cli}/index.js

@@ -80,9 +80,9 @@ var COMMANDS = {
             return /*#__PURE__*/ _interop_require_wildcard(require("./query.js"));
         });
     },
-    find: function find() {
+    search: function search() {
         return Promise.resolve().then(function() {
-            return /*#__PURE__*/ _interop_require_wildcard(require("./find.js"));
+            return /*#__PURE__*/ _interop_require_wildcard(require("./search.js"));
         });
     },
     map: function map() {

package/dist/cjs/cli/index.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/cli/index.ts"],"sourcesContent":["import type { Command } from './types.ts';\n\n// Lazy registry: a command's imports load only when it runs, so a heavy dependency in one\n// command never taxes the others (or --version/--help). These names are reserved: a named\n// query in config can never shadow them.\nexport const COMMANDS: Record<string, () => Promise<{ default: Command }>> = {\n  init: () => import('./init.ts'),\n  watch: () => import('./watch.ts'),\n  status: () => import('./status.ts'),\n  check: () => import('./check.ts'),\n  rebuild: () => import('./rebuild.ts'),\n  query: () => import('./query.ts'),\n  search: () => import('./search.ts'),\n  map: () => import('./map.ts'),\n  peek: () => import('./peek.ts'),\n};\n\nexport type { Command, Ctx } from './types.ts';\n"],"names":["COMMANDS","init","watch","status","check","rebuild","query","search","map","peek"],"mappings":";;;;+BAKaA;;;eAAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAAN,IAAMA,WAAgE;IAC3EC,MAAM,SAANA;eAAY;2DAAA,QAAO;;;IACnBC,OAAO,SAAPA;eAAa;2DAAA,QAAO;;;IACpBC,QAAQ,SAARA;eAAc;2DAAA,QAAO;;;IACrBC,OAAO,SAAPA;eAAa;2DAAA,QAAO;;;IACpBC,SAAS,SAATA;eAAe;2DAAA,QAAO;;;IACtBC,OAAO,SAAPA;eAAa;2DAAA,QAAO;;;IACpBC,QAAQ,SAARA;eAAc;2DAAA,QAAO;;;IACrBC,KAAK,SAALA;eAAW;2DAAA,QAAO;;;IAClBC,MAAM,SAANA;eAAY;2DAAA,QAAO;;;AACrB"}

package/dist/cjs/{commands → cli}/init.js

@@ -13,6 +13,9 @@ var init = function init(ctx) {
     var configPath = (0, _configts.initConfig)(process.cwd());
     console.log("created ".concat(configPath));
     console.log("query away: ".concat(ctx.name, ' query "SELECT path FROM frontmatter LIMIT 10"'));
+    // The decisions a new tree faces -- features, frontmatter conventions, note size -- are not
+    // in this output; name both places that hold them, since an agent may have neither.
+    console.log('features and tree design: the sense-setup skill, or schema.json');
 };
 var _default = init;
 /* CJS INTEROP */ if (exports.__esModule && exports.default) { try { Object.defineProperty(exports.default, '__esModule', { value: true }); for (var key in exports) { exports.default[key] = exports[key]; } } catch (_) {}; module.exports = exports.default; }

package/dist/cjs/cli/init.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/cli/init.ts"],"sourcesContent":["import { initConfig } from '../config.ts';\nimport type { Command } from './types.ts';\n\nconst init: Command = (ctx) => {\n  const configPath = initConfig(process.cwd());\n  console.log(`created ${configPath}`);\n  console.log(`query away: ${ctx.name} query \"SELECT path FROM frontmatter LIMIT 10\"`);\n  // The decisions a new tree faces -- features, frontmatter conventions, note size -- are not\n  // in this output; name both places that hold them, since an agent may have neither.\n  console.log('features and tree design: the sense-setup skill, or schema.json');\n};\nexport default init;\n"],"names":["init","ctx","configPath","initConfig","process","cwd","console","log","name"],"mappings":";;;;+BAWA;;;eAAA;;;wBAX2B;AAG3B,IAAMA,OAAgB,cAACC;IACrB,IAAMC,aAAaC,IAAAA,oBAAU,EAACC,QAAQC,GAAG;IACzCC,QAAQC,GAAG,CAAC,AAAC,WAAqB,OAAXL;IACvBI,QAAQC,GAAG,CAAC,AAAC,eAAuB,OAATN,IAAIO,IAAI,EAAC;IACpC,4FAA4F;IAC5F,oFAAoF;IACpFF,QAAQC,GAAG,CAAC;AACd;IACA,WAAeP"}