@trazum/core 1.8.0

package/dist/extract.js

@@ -0,0 +1,280 @@
+/**
+ * Prompts embedded in source files.
+ *
+ * `check` and `diff` read `.txt`, `.md`, `.prompt` and `.tmpl`. Real prompts live
+ * in TypeScript template literals, Python triple-quoted strings and YAML blocks,
+ * so adopting Trazum has meant first refactoring them out into standalone files —
+ * a change to somebody's application as the price of admission.
+ *
+ * **It reads a marker, it does not guess.** `// trazum:prompt` on the line before
+ * the literal, and then the literal by delimiter matching. Guessing which string
+ * in a file is a prompt is a heuristic, and a heuristic inside a tool used as a
+ * CI gate produces failures on strings nobody meant to govern. A marker is one
+ * line of noise in exchange for never being wrong about what it picked up.
+ *
+ * **Interpolation already works and is not a special case.** `${x}` inside a
+ * template literal is exactly the placeholder shape `segment.ts` protects, so an
+ * embedded prompt gets the same cache-prefix analysis, the same protection from
+ * the rules, and the same `--reorder` treatment as a `{{x}}` template. Nothing
+ * here has to know about it.
+ *
+ * **The honest limit:** a prompt assembled from concatenated pieces cannot be
+ * read this way. `` `You are ${role}.` + rules.join('\n') `` is a prompt whose
+ * text does not exist until it runs. This module declines it and says so rather
+ * than governing the half it can see — a budget enforced against a fragment is a
+ * green build for a prompt nobody measured.
+ *
+ * Scanned character by character rather than with a regex. The module it most
+ * resembles shipped two quadratic patterns this week, and delimiter matching over
+ * untrusted source is exactly the shape that goes wrong.
+ */
+/** The marker, in the comment syntaxes that cover the languages prompts live in. */
+const MARKERS = ['//', '#', '--', '<!--'];
+const TAG = 'trazum:prompt';
+/**
+ * Both ways an HTML comment can close.
+ *
+ * `--!>` is the "comment end bang" the HTML parser also accepts, and stripping
+ * only `-->` put it in the name: `<!-- trazum:prompt greeting--!>` produced the
+ * name `greeting--!>`. Found by CodeQL, which is the second time this week a
+ * pattern has been right about the case in front of it and wrong about the one
+ * beside it.
+ */
+const COMMENT_CLOSE = /--!?>\s*$/;
+/**
+ * What a name may be.
+ *
+ * An identifier charset rather than "whatever is left on the line". The name
+ * flows into `promptId`, which is printed in reports and matched against the
+ * budget patterns in `trazum.config.json` — a name is an identifier, and letting
+ * it be arbitrary text is how a stray terminator became part of one above. A
+ * candidate that does not fit falls back to the `file:line` form, which always
+ * works.
+ */
+const NAME = /^[A-Za-z0-9][A-Za-z0-9._-]*$/;
+/**
+ * How far past the marker a literal may start.
+ *
+ * A prompt is expected on the next line or two — `const SYSTEM = \`` and the
+ * occasional type annotation. Scanning further would let a marker attach itself
+ * to an unrelated string much later in the file, which is the false positive this
+ * design exists to avoid.
+ */
+const MAX_GAP_LINES = 3;
+/**
+ * A forward-only line counter.
+ *
+ * Markers are found in increasing order, so each character is counted once
+ * across the whole scan. Recomputing the line by counting from zero at every
+ * marker is the obvious version and is quadratic in the number of markers —
+ * 15.5 seconds on a file of 20,000 of them, which the hostile-input tests
+ * caught before this shipped.
+ */
+function lineCounter(source) {
+    let seen = 0;
+    let line = 1;
+    return (offset) => {
+        for (; seen < offset && seen < source.length; seen++) {
+            if (source[seen] === '\n')
+                line++;
+        }
+        return line;
+    };
+}
+const ESCAPES = {
+    n: '\n',
+    t: '\t',
+    r: '\r',
+    '\\': '\\',
+    '"': '"',
+    "'": "'",
+    '`': '`',
+    '0': '\0',
+};
+/**
+ * Reads a string literal starting at `start`, which must be its opening quote.
+ *
+ * Returns the resolved text and the offset just past the closing delimiter, or
+ * null when the literal never closes. Escapes are resolved for single- and
+ * double-quoted strings, where `\n` is one newline written as two characters;
+ * backtick and triple-quoted literals carry their newlines literally and only
+ * need the delimiter itself unescaped.
+ */
+function readLiteral(source, start) {
+    const triple = source.slice(start, start + 3);
+    const quote = triple === '"""' || triple === "'''"
+        ? triple
+        : source[start] === '`' || source[start] === '"' || source[start] === "'"
+            ? source[start]
+            : null;
+    if (quote === null)
+        return null;
+    const raw = quote === '`' || quote.length === 3;
+    const textStart = start + quote.length;
+    let out = '';
+    let i = textStart;
+    while (i < source.length) {
+        const ch = source[i];
+        if (ch === '\\') {
+            const next = source[i + 1];
+            if (next === undefined)
+                break;
+            // In a raw literal only the delimiter and the backslash itself are
+            // escapes; `\n` there is a literal backslash followed by an n, and
+            // resolving it would invent a newline the prompt does not contain.
+            if (raw) {
+                out += next === quote[0] || next === '\\' ? next : `\\${next}`;
+            }
+            else {
+                out += ESCAPES[next] ?? `\\${next}`;
+            }
+            i += 2;
+            continue;
+        }
+        if (source.startsWith(quote, i)) {
+            return { text: out, quote, end: i + quote.length, textStart };
+        }
+        // A single-quoted or double-quoted literal cannot span lines. Stopping here
+        // rather than running to the end of the file is what keeps a missing quote
+        // from swallowing the rest of the source.
+        if (quote.length === 1 && quote !== '`' && ch === '\n')
+            return null;
+        out += ch;
+        i++;
+    }
+    return null;
+}
+/** Whether the expression continues past the literal with a concatenation. */
+function isConcatenated(source, end) {
+    let i = end;
+    while (i < source.length && (source[i] === ' ' || source[i] === '\t'))
+        i++;
+    // `+` in JS/TS, `.` in PHP-ish sources, and a bare newline followed by `+` is
+    // the wrapped form of the same thing. A trailing comma or semicolon is not.
+    if (source[i] === '+')
+        return true;
+    if (source[i] === '\n') {
+        let j = i + 1;
+        while (j < source.length && /[ \t\n]/.test(source[j]))
+            j++;
+        return source[j] === '+';
+    }
+    return false;
+}
+/**
+ * Finds every marked prompt in a source file.
+ *
+ * Returns what it found and what it refused, because a marker that produced
+ * nothing is the case the author most needs to hear about: they asked for the
+ * prompt to be governed and it is not being governed.
+ */
+export function extractPrompts(source) {
+    const prompts = [];
+    const declined = [];
+    const lineAt = lineCounter(source);
+    let cursor = 0;
+    while (cursor < source.length) {
+        const at = source.indexOf(TAG, cursor);
+        if (at === -1)
+            break;
+        cursor = at + TAG.length;
+        // The tag has to be inside a comment. Without this, the string 'trazum:prompt'
+        // appearing in a prompt's own text would mark the prompt after it.
+        const lineStart = source.lastIndexOf('\n', at) + 1;
+        const before = source.slice(lineStart, at).trimStart();
+        if (!MARKERS.some((m) => before.startsWith(m)))
+            continue;
+        const markerLine = lineAt(at);
+        // The rest of the marker line is an optional name: `trazum:prompt greeting`.
+        const lineEnd = source.indexOf('\n', at);
+        const rest = source.slice(cursor, lineEnd === -1 ? source.length : lineEnd);
+        const candidate = rest.replace(COMMENT_CLOSE, '').trim().split(/\s+/)[0] ?? '';
+        const name = NAME.test(candidate) ? candidate : undefined;
+        // Scan forward for the opening delimiter, bounded so a marker cannot adopt
+        // a string much further down the file.
+        let i = lineEnd === -1 ? source.length : lineEnd + 1;
+        let lines = 0;
+        let opened = null;
+        let openAt = -1;
+        while (i < source.length && lines <= MAX_GAP_LINES) {
+            const ch = source[i];
+            if (ch === '\n') {
+                lines++;
+                i++;
+                continue;
+            }
+            if (ch === '`' || ch === '"' || ch === "'") {
+                openAt = i;
+                opened = readLiteral(source, i);
+                break;
+            }
+            i++;
+        }
+        if (openAt === -1) {
+            declined.push({
+                line: markerLine,
+                reason: 'no-literal',
+                detail: `no string literal within ${MAX_GAP_LINES} lines of the marker`,
+            });
+            continue;
+        }
+        if (opened === null) {
+            declined.push({
+                line: markerLine,
+                reason: 'unterminated',
+                detail: 'the literal is never closed',
+            });
+            continue;
+        }
+        if (isConcatenated(source, opened.end)) {
+            declined.push({
+                line: markerLine,
+                reason: 'concatenated',
+                detail: 'the prompt is built by concatenation, so its text does not exist until it runs',
+            });
+            cursor = opened.end;
+            continue;
+        }
+        prompts.push({
+            name,
+            line: markerLine,
+            text: opened.text,
+            quote: opened.quote,
+            start: opened.textStart,
+            end: opened.end - opened.quote.length,
+        });
+        cursor = opened.end;
+    }
+    return { prompts, declined };
+}
+/**
+ * A stable identifier for an embedded prompt, for budgets and reports.
+ *
+ * `src/prompts.ts#support` when the marker was named, `src/prompts.ts:12` when it
+ * was not. Both are path-prefixed so the glob patterns in `trazum.config.json`
+ * cover embedded prompts without learning a new syntax — `src/**` matches either.
+ */
+export function promptId(path, prompt) {
+    return prompt.name ? `${path}#${prompt.name}` : `${path}:${prompt.line}`;
+}
+/** Whether a file is worth opening for extraction at all. */
+export const SOURCE_EXTENSIONS = [
+    '.ts',
+    '.tsx',
+    '.js',
+    '.jsx',
+    '.mjs',
+    '.py',
+    '.rb',
+    '.go',
+    '.rs',
+    '.java',
+    '.kt',
+    '.cs',
+    '.php',
+    '.yaml',
+    '.yml',
+];
+/** Cheap pre-filter: a source file with no marker has nothing to extract. */
+export const hasMarker = (source) => source.includes(TAG);
+//# sourceMappingURL=extract.js.map

package/dist/extract.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"extract.js","sourceRoot":"","sources":["../src/extract.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA6BG;AAmCH,oFAAoF;AACpF,MAAM,OAAO,GAAG,CAAC,IAAI,EAAE,GAAG,EAAE,IAAI,EAAE,MAAM,CAAC,CAAC;AAC1C,MAAM,GAAG,GAAG,eAAe,CAAC;AAE5B;;;;;;;;GAQG;AACH,MAAM,aAAa,GAAG,WAAW,CAAC;AAElC;;;;;;;;;GASG;AACH,MAAM,IAAI,GAAG,8BAA8B,CAAC;AAE5C;;;;;;;GAOG;AACH,MAAM,aAAa,GAAG,CAAC,CAAC;AAExB;;;;;;;;GAQG;AACH,SAAS,WAAW,CAAC,MAAc;IACjC,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,OAAO,CAAC,MAAM,EAAE,EAAE;QAChB,OAAO,IAAI,GAAG,MAAM,IAAI,IAAI,GAAG,MAAM,CAAC,MAAM,EAAE,IAAI,EAAE,EAAE,CAAC;YACrD,IAAI,MAAM,CAAC,IAAI,CAAC,KAAK,IAAI;gBAAE,IAAI,EAAE,CAAC;QACpC,CAAC;QACD,OAAO,IAAI,CAAC;IACd,CAAC,CAAC;AACJ,CAAC;AAED,MAAM,OAAO,GAA2B;IACtC,CAAC,EAAE,IAAI;IACP,CAAC,EAAE,IAAI;IACP,CAAC,EAAE,IAAI;IACP,IAAI,EAAE,IAAI;IACV,GAAG,EAAE,GAAG;IACR,GAAG,EAAE,GAAG;IACR,GAAG,EAAE,GAAG;IACR,GAAG,EAAE,IAAI;CACV,CAAC;AAEF;;;;;;;;GAQG;AACH,SAAS,WAAW,CAClB,MAAc,EACd,KAAa;IAEb,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,KAAK,EAAE,KAAK,GAAG,CAAC,CAAC,CAAC;IAC9C,MAAM,KAAK,GACT,MAAM,KAAK,KAAK,IAAI,MAAM,KAAK,KAAK;QAClC,CAAC,CAAE,MAAwB;QAC3B,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,KAAK,GAAG,IAAI,MAAM,CAAC,KAAK,CAAC,KAAK,GAAG,IAAI,MAAM,CAAC,KAAK,CAAC,KAAK,GAAG;YACvE,CAAC,CAAE,MAAM,CAAC,KAAK,CAAqB;YACpC,CAAC,CAAC,IAAI,CAAC;IACb,IAAI,KAAK,KAAK,IAAI;QAAE,OAAO,IAAI,CAAC;IAEhC,MAAM,GAAG,GAAG,KAAK,KAAK,GAAG,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC,CAAC;IAChD,MAAM,SAAS,GAAG,KAAK,GAAG,KAAK,CAAC,MAAM,CAAC;IACvC,IAAI,GAAG,GAAG,EAAE,CAAC;IACb,IAAI,CAAC,GAAG,SAAS,CAAC;IAElB,OAAO,CAAC,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC;QACzB,MAAM,EAAE,GAAG,MAAM,CAAC,CAAC,CAAE,CAAC;QAEtB,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC;YAChB,MAAM,IAAI,GAAG,MAAM,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;YAC3B,IAAI,IAAI,KAAK,SAAS;gBAAE,MAAM;YAC9B,mEAAmE;YACnE,mEAAmE;YACnE,mEAAmE;YACnE,IAAI,GAAG,EAAE,CAAC;gBACR,GAAG,IAAI,IAAI,KAAK,KAAK,CAAC,CAAC,CAAC,IAAI,IAAI,KAAK,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YACjE,CAAC;iBAAM,CAAC;gBACN,GAAG,IAAI,OAAO,CAAC,IAAI,CAAC,IAAI,KAAK,IAAI,EAAE,CAAC;YACtC,CAAC;YACD,CAAC,IAAI,CAAC,CAAC;YACP,SAAS;QACX,CAAC;QAED,IAAI,MAAM,CAAC,UAAU,CAAC,KAAK,EAAE,CAAC,CAAC,EAAE,CAAC;YAChC,OAAO,EAAE,IAAI,EAAE,GAAG,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,SAAS,EAAE,CAAC;QAChE,CAAC;QAED,4EAA4E;QAC5E,2EAA2E;QAC3E,0CAA0C;QAC1C,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC,IAAI,KAAK,KAAK,GAAG,IAAI,EAAE,KAAK,IAAI;YAAE,OAAO,IAAI,CAAC;QAEpE,GAAG,IAAI,EAAE,CAAC;QACV,CAAC,EAAE,CAAC;IACN,CAAC;IAED,OAAO,IAAI,CAAC;AACd,CAAC;AAED,8EAA8E;AAC9E,SAAS,cAAc,CAAC,MAAc,EAAE,GAAW;IACjD,IAAI,CAAC,GAAG,GAAG,CAAC;IACZ,OAAO,CAAC,GAAG,MAAM,CAAC,MAAM,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,KAAK,GAAG,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,CAAC;QAAE,CAAC,EAAE,CAAC;IAC3E,8EAA8E;IAC9E,4EAA4E;IAC5E,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,GAAG;QAAE,OAAO,IAAI,CAAC;IACnC,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACvB,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;QACd,OAAO,CAAC,GAAG,MAAM,CAAC,MAAM,IAAI,SAAS,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAE,CAAC;YAAE,CAAC,EAAE,CAAC;QAC5D,OAAO,MAAM,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC;IAC3B,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED;;;;;;GAMG;AACH,MAAM,UAAU,cAAc,CAAC,MAAc;IAC3C,MAAM,OAAO,GAAsB,EAAE,CAAC;IACtC,MAAM,QAAQ,GAAqB,EAAE,CAAC;IAEtC,MAAM,MAAM,GAAG,WAAW,CAAC,MAAM,CAAC,CAAC;IAEnC,IAAI,MAAM,GAAG,CAAC,CAAC;IACf,OAAO,MAAM,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC;QAC9B,MAAM,EAAE,GAAG,MAAM,CAAC,OAAO,CAAC,GAAG,EAAE,MAAM,CAAC,CAAC;QACvC,IAAI,EAAE,KAAK,CAAC,CAAC;YAAE,MAAM;QACrB,MAAM,GAAG,EAAE,GAAG,GAAG,CAAC,MAAM,CAAC;QAEzB,+EAA+E;QAC/E,mEAAmE;QACnE,MAAM,SAAS,GAAG,MAAM,CAAC,WAAW,CAAC,IAAI,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC;QACnD,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,SAAS,EAAE,EAAE,CAAC,CAAC,SAAS,EAAE,CAAC;QACvD,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,MAAM,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC;YAAE,SAAS;QAEzD,MAAM,UAAU,GAAG,MAAM,CAAC,EAAE,CAAC,CAAC;QAE9B,6EAA6E;QAC7E,MAAM,OAAO,GAAG,MAAM,CAAC,OAAO,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC;QACzC,MAAM,IAAI,GAAG,MAAM,CAAC,KAAK,CAAC,MAAM,EAAE,OAAO,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC;QAC5E,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,aAAa,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAC/E,MAAM,IAAI,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,SAAS,CAAC;QAE1D,2EAA2E;QAC3E,uCAAuC;QACvC,IAAI,CAAC,GAAG,OAAO,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,GAAG,CAAC,CAAC;QACrD,IAAI,KAAK,GAAG,CAAC,CAAC;QACd,IAAI,MAAM,GAAmC,IAAI,CAAC;QAClD,IAAI,MAAM,GAAG,CAAC,CAAC,CAAC;QAEhB,OAAO,CAAC,GAAG,MAAM,CAAC,MAAM,IAAI,KAAK,IAAI,aAAa,EAAE,CAAC;YACnD,MAAM,EAAE,GAAG,MAAM,CAAC,CAAC,CAAE,CAAC;YACtB,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC;gBAChB,KAAK,EAAE,CAAC;gBACR,CAAC,EAAE,CAAC;gBACJ,SAAS;YACX,CAAC;YACD,IAAI,EAAE,KAAK,GAAG,IAAI,EAAE,KAAK,GAAG,IAAI,EAAE,KAAK,GAAG,EAAE,CAAC;gBAC3C,MAAM,GAAG,CAAC,CAAC;gBACX,MAAM,GAAG,WAAW,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC;gBAChC,MAAM;YACR,CAAC;YACD,CAAC,EAAE,CAAC;QACN,CAAC;QAED,IAAI,MAAM,KAAK,CAAC,CAAC,EAAE,CAAC;YAClB,QAAQ,CAAC,IAAI,CAAC;gBACZ,IAAI,EAAE,UAAU;gBAChB,MAAM,EAAE,YAAY;gBACpB,MAAM,EAAE,4BAA4B,aAAa,sBAAsB;aACxE,CAAC,CAAC;YACH,SAAS;QACX,CAAC;QACD,IAAI,MAAM,KAAK,IAAI,EAAE,CAAC;YACpB,QAAQ,CAAC,IAAI,CAAC;gBACZ,IAAI,EAAE,UAAU;gBAChB,MAAM,EAAE,cAAc;gBACtB,MAAM,EAAE,6BAA6B;aACtC,CAAC,CAAC;YACH,SAAS;QACX,CAAC;QACD,IAAI,cAAc,CAAC,MAAM,EAAE,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC;YACvC,QAAQ,CAAC,IAAI,CAAC;gBACZ,IAAI,EAAE,UAAU;gBAChB,MAAM,EAAE,cAAc;gBACtB,MAAM,EACJ,gFAAgF;aACnF,CAAC,CAAC;YACH,MAAM,GAAG,MAAM,CAAC,GAAG,CAAC;YACpB,SAAS;QACX,CAAC;QAED,OAAO,CAAC,IAAI,CAAC;YACX,IAAI;YACJ,IAAI,EAAE,UAAU;YAChB,IAAI,EAAE,MAAM,CAAC,IAAI;YACjB,KAAK,EAAE,MAAM,CAAC,KAAK;YACnB,KAAK,EAAE,MAAM,CAAC,SAAS;YACvB,GAAG,EAAE,MAAM,CAAC,GAAG,GAAG,MAAM,CAAC,KAAK,CAAC,MAAM;SACtC,CAAC,CAAC;QACH,MAAM,GAAG,MAAM,CAAC,GAAG,CAAC;IACtB,CAAC;IAED,OAAO,EAAE,OAAO,EAAE,QAAQ,EAAE,CAAC;AAC/B,CAAC;AAED;;;;;;GAMG;AACH,MAAM,UAAU,QAAQ,CAAC,IAAY,EAAE,MAAuB;IAC5D,OAAO,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,IAAI,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,GAAG,IAAI,IAAI,MAAM,CAAC,IAAI,EAAE,CAAC;AAC3E,CAAC;AAED,6DAA6D;AAC7D,MAAM,CAAC,MAAM,iBAAiB,GAAG;IAC/B,KAAK;IACL,MAAM;IACN,KAAK;IACL,MAAM;IACN,MAAM;IACN,KAAK;IACL,KAAK;IACL,KAAK;IACL,KAAK;IACL,OAAO;IACP,KAAK;IACL,KAAK;IACL,MAAM;IACN,OAAO;IACP,MAAM;CACP,CAAC;AAEF,6EAA6E;AAC7E,MAAM,CAAC,MAAM,SAAS,GAAG,CAAC,MAAc,EAAW,EAAE,CAAC,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC"}

package/dist/gcp-auth.d.ts

@@ -0,0 +1,58 @@
+/**
+ * A Google service-account access token, without the SDK.
+ *
+ * Vertex will not take an API key. It wants an OAuth access token, and the only
+ * way to get one unattended is the two-legged flow: build a JWT, sign it with
+ * the service account's private key, and trade it at the token endpoint.
+ * `google-auth-library` is ninety-odd packages to do that; this library has zero
+ * runtime dependencies and a test that fails the build if one appears, because
+ * every dependency is somebody else's code reading your prompts.
+ *
+ * WebCrypto rather than `node:crypto`, so the browser-safe entry point stays
+ * browser-safe. `RSASSA-PKCS1-v1_5` with SHA-256 is what Google calls RS256.
+ */
+/**
+ * PEM to the DER bytes `importKey` wants.
+ *
+ * The header, the footer and every newline come out. A PEM that still has them
+ * fails inside WebCrypto with `DataError`, which says nothing about which of the
+ * fourteen things that can be wrong with a key is wrong with this one.
+ */
+export declare function pkcs8FromPem(pem: string): ArrayBuffer;
+export interface ServiceAccount {
+    client_email: string;
+    private_key: string;
+    token_uri?: string;
+}
+/**
+ * The assertion Google trades for a token.
+ *
+ * Exported so a test can read it: this is the document where a wrong `aud`, a
+ * clock an hour out, or a scope nobody granted turns into `invalid_grant`, an
+ * error message that names none of the three.
+ */
+export declare function signedJwt(account: ServiceAccount, scope: string, now: Date): Promise<string>;
+export interface CachedToken {
+    token: string;
+    /** Epoch seconds. */
+    expiresAt: number;
+}
+/**
+ * Trades the assertion for an access token, caching until shortly before expiry.
+ *
+ * The cache is the point. A token lasts an hour and `optimize --suggest` over a
+ * directory makes one call per prompt: fetching a token each time turns forty
+ * prompts into eighty requests, half of them to an endpoint that rate-limits.
+ *
+ * Sixty seconds of margin, because a token that expires in flight fails the
+ * request it was fetched for, and the clock here is not the clock there.
+ */
+export declare function accessToken(account: ServiceAccount, options?: {
+    scope?: string;
+    fetchImpl?: typeof fetch;
+    now?: () => Date;
+    cache?: {
+        current: CachedToken | null;
+    };
+}): Promise<string>;
+//# sourceMappingURL=gcp-auth.d.ts.map

package/dist/gcp-auth.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"gcp-auth.d.ts","sourceRoot":"","sources":["../src/gcp-auth.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AAaH;;;;;;GAMG;AACH,wBAAgB,YAAY,CAAC,GAAG,EAAE,MAAM,GAAG,WAAW,CAiBrD;AAED,MAAM,WAAW,cAAc;IAC7B,YAAY,EAAE,MAAM,CAAC;IACrB,WAAW,EAAE,MAAM,CAAC;IACpB,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED;;;;;;GAMG;AACH,wBAAsB,SAAS,CAC7B,OAAO,EAAE,cAAc,EACvB,KAAK,EAAE,MAAM,EACb,GAAG,EAAE,IAAI,GACR,OAAO,CAAC,MAAM,CAAC,CA8BjB;AAED,MAAM,WAAW,WAAW;IAC1B,KAAK,EAAE,MAAM,CAAC;IACd,qBAAqB;IACrB,SAAS,EAAE,MAAM,CAAC;CACnB;AAED;;;;;;;;;GASG;AACH,wBAAsB,WAAW,CAC/B,OAAO,EAAE,cAAc,EACvB,OAAO,GAAE;IACP,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,SAAS,CAAC,EAAE,OAAO,KAAK,CAAC;IACzB,GAAG,CAAC,EAAE,MAAM,IAAI,CAAC;IACjB,KAAK,CAAC,EAAE;QAAE,OAAO,EAAE,WAAW,GAAG,IAAI,CAAA;KAAE,CAAC;CACpC,GACL,OAAO,CAAC,MAAM,CAAC,CAyCjB"}

package/dist/gcp-auth.js

@@ -0,0 +1,113 @@
+/**
+ * A Google service-account access token, without the SDK.
+ *
+ * Vertex will not take an API key. It wants an OAuth access token, and the only
+ * way to get one unattended is the two-legged flow: build a JWT, sign it with
+ * the service account's private key, and trade it at the token endpoint.
+ * `google-auth-library` is ninety-odd packages to do that; this library has zero
+ * runtime dependencies and a test that fails the build if one appears, because
+ * every dependency is somebody else's code reading your prompts.
+ *
+ * WebCrypto rather than `node:crypto`, so the browser-safe entry point stays
+ * browser-safe. `RSASSA-PKCS1-v1_5` with SHA-256 is what Google calls RS256.
+ */
+/** Base64url, no padding — what a JWT uses and what `btoa` does not produce. */
+function base64url(bytes) {
+    let binary = '';
+    for (const byte of bytes)
+        binary += String.fromCharCode(byte);
+    return btoa(binary).replace(/\+/g, '-').replace(/\//g, '_').replace(/=+$/, '');
+}
+const encoder = new TextEncoder();
+const encodeJson = (value) => base64url(encoder.encode(JSON.stringify(value)));
+/**
+ * PEM to the DER bytes `importKey` wants.
+ *
+ * The header, the footer and every newline come out. A PEM that still has them
+ * fails inside WebCrypto with `DataError`, which says nothing about which of the
+ * fourteen things that can be wrong with a key is wrong with this one.
+ */
+export function pkcs8FromPem(pem) {
+    const body = pem
+        .replace(/-----BEGIN [^-]+-----/, '')
+        .replace(/-----END [^-]+-----/, '')
+        .replace(/\s+/g, '');
+    if (!body)
+        throw new Error('The service account private key is empty.');
+    let binary;
+    try {
+        binary = atob(body);
+    }
+    catch {
+        throw new Error('The service account private key is not valid base64.');
+    }
+    // Returned as an ArrayBuffer rather than a view: `importKey` accepts a
+    // `BufferSource`, and a `Uint8Array` over a `SharedArrayBuffer` is not one
+    // as far as the type checker is concerned.
+    return Uint8Array.from(binary, (character) => character.charCodeAt(0)).buffer;
+}
+/**
+ * The assertion Google trades for a token.
+ *
+ * Exported so a test can read it: this is the document where a wrong `aud`, a
+ * clock an hour out, or a scope nobody granted turns into `invalid_grant`, an
+ * error message that names none of the three.
+ */
+export async function signedJwt(account, scope, now) {
+    const tokenUri = account.token_uri ?? 'https://oauth2.googleapis.com/token';
+    const issued = Math.floor(now.getTime() / 1000);
+    const header = encodeJson({ alg: 'RS256', typ: 'JWT' });
+    const claims = encodeJson({
+        iss: account.client_email,
+        scope,
+        aud: tokenUri,
+        iat: issued,
+        // An hour is Google's maximum. Longer is rejected outright rather than
+        // clamped, which is a confusing way to learn about a limit.
+        exp: issued + 3600,
+    });
+    const key = await crypto.subtle.importKey('pkcs8', pkcs8FromPem(account.private_key), { name: 'RSASSA-PKCS1-v1_5', hash: 'SHA-256' }, false, ['sign']);
+    const signature = await crypto.subtle.sign('RSASSA-PKCS1-v1_5', key, encoder.encode(`${header}.${claims}`));
+    return `${header}.${claims}.${base64url(new Uint8Array(signature))}`;
+}
+/**
+ * Trades the assertion for an access token, caching until shortly before expiry.
+ *
+ * The cache is the point. A token lasts an hour and `optimize --suggest` over a
+ * directory makes one call per prompt: fetching a token each time turns forty
+ * prompts into eighty requests, half of them to an endpoint that rate-limits.
+ *
+ * Sixty seconds of margin, because a token that expires in flight fails the
+ * request it was fetched for, and the clock here is not the clock there.
+ */
+export async function accessToken(account, options = {}) {
+    const { scope = 'https://www.googleapis.com/auth/cloud-platform', fetchImpl = fetch, now = () => new Date(), cache, } = options;
+    const at = now();
+    const seconds = Math.floor(at.getTime() / 1000);
+    if (cache?.current && cache.current.expiresAt - 60 > seconds) {
+        return cache.current.token;
+    }
+    const assertion = await signedJwt(account, scope, at);
+    const tokenUri = account.token_uri ?? 'https://oauth2.googleapis.com/token';
+    const res = await fetchImpl(tokenUri, {
+        method: 'POST',
+        redirect: 'error',
+        headers: { 'content-type': 'application/x-www-form-urlencoded' },
+        body: new URLSearchParams({
+            grant_type: 'urn:ietf:params:oauth:grant-type:jwt-bearer',
+            assertion,
+        }).toString(),
+    });
+    if (!res.ok) {
+        throw new Error(`Google refused the service account assertion (${res.status}): ${await res.text()}`);
+    }
+    const data = (await res.json());
+    if (typeof data.access_token !== 'string' || !data.access_token) {
+        throw new Error('Google returned no access_token for the service account.');
+    }
+    const lifetime = typeof data.expires_in === 'number' ? data.expires_in : 3600;
+    if (cache)
+        cache.current = { token: data.access_token, expiresAt: seconds + lifetime };
+    return data.access_token;
+}
+//# sourceMappingURL=gcp-auth.js.map

package/dist/gcp-auth.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"gcp-auth.js","sourceRoot":"","sources":["../src/gcp-auth.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AAEH,gFAAgF;AAChF,SAAS,SAAS,CAAC,KAAiB;IAClC,IAAI,MAAM,GAAG,EAAE,CAAC;IAChB,KAAK,MAAM,IAAI,IAAI,KAAK;QAAE,MAAM,IAAI,MAAM,CAAC,YAAY,CAAC,IAAI,CAAC,CAAC;IAC9D,OAAO,IAAI,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC;AACjF,CAAC;AAED,MAAM,OAAO,GAAG,IAAI,WAAW,EAAE,CAAC;AAElC,MAAM,UAAU,GAAG,CAAC,KAAc,EAAU,EAAE,CAAC,SAAS,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,SAAS,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;AAEhG;;;;;;GAMG;AACH,MAAM,UAAU,YAAY,CAAC,GAAW;IACtC,MAAM,IAAI,GAAG,GAAG;SACb,OAAO,CAAC,uBAAuB,EAAE,EAAE,CAAC;SACpC,OAAO,CAAC,qBAAqB,EAAE,EAAE,CAAC;SAClC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC;IACvB,IAAI,CAAC,IAAI;QAAE,MAAM,IAAI,KAAK,CAAC,2CAA2C,CAAC,CAAC;IAExE,IAAI,MAAc,CAAC;IACnB,IAAI,CAAC;QACH,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC;IACtB,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,IAAI,KAAK,CAAC,sDAAsD,CAAC,CAAC;IAC1E,CAAC;IACD,uEAAuE;IACvE,2EAA2E;IAC3E,2CAA2C;IAC3C,OAAO,UAAU,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC,SAAS,EAAE,EAAE,CAAC,SAAS,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC,MAAqB,CAAC;AAC/F,CAAC;AAQD;;;;;;GAMG;AACH,MAAM,CAAC,KAAK,UAAU,SAAS,CAC7B,OAAuB,EACvB,KAAa,EACb,GAAS;IAET,MAAM,QAAQ,GAAG,OAAO,CAAC,SAAS,IAAI,qCAAqC,CAAC;IAC5E,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,OAAO,EAAE,GAAG,IAAI,CAAC,CAAC;IAEhD,MAAM,MAAM,GAAG,UAAU,CAAC,EAAE,GAAG,EAAE,OAAO,EAAE,GAAG,EAAE,KAAK,EAAE,CAAC,CAAC;IACxD,MAAM,MAAM,GAAG,UAAU,CAAC;QACxB,GAAG,EAAE,OAAO,CAAC,YAAY;QACzB,KAAK;QACL,GAAG,EAAE,QAAQ;QACb,GAAG,EAAE,MAAM;QACX,uEAAuE;QACvE,4DAA4D;QAC5D,GAAG,EAAE,MAAM,GAAG,IAAI;KACnB,CAAC,CAAC;IAEH,MAAM,GAAG,GAAG,MAAM,MAAM,CAAC,MAAM,CAAC,SAAS,CACvC,OAAO,EACP,YAAY,CAAC,OAAO,CAAC,WAAW,CAAC,EACjC,EAAE,IAAI,EAAE,mBAAmB,EAAE,IAAI,EAAE,SAAS,EAAE,EAC9C,KAAK,EACL,CAAC,MAAM,CAAC,CACT,CAAC;IAEF,MAAM,SAAS,GAAG,MAAM,MAAM,CAAC,MAAM,CAAC,IAAI,CACxC,mBAAmB,EACnB,GAAG,EACH,OAAO,CAAC,MAAM,CAAC,GAAG,MAAM,IAAI,MAAM,EAAE,CAAC,CACtC,CAAC;IAEF,OAAO,GAAG,MAAM,IAAI,MAAM,IAAI,SAAS,CAAC,IAAI,UAAU,CAAC,SAAS,CAAC,CAAC,EAAE,CAAC;AACvE,CAAC;AAQD;;;;;;;;;GASG;AACH,MAAM,CAAC,KAAK,UAAU,WAAW,CAC/B,OAAuB,EACvB,OAAO,GAKH,EAAE;IAEN,MAAM,EACJ,KAAK,GAAG,gDAAgD,EACxD,SAAS,GAAG,KAAK,EACjB,GAAG,GAAG,GAAG,EAAE,CAAC,IAAI,IAAI,EAAE,EACtB,KAAK,GACN,GAAG,OAAO,CAAC;IAEZ,MAAM,EAAE,GAAG,GAAG,EAAE,CAAC;IACjB,MAAM,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,OAAO,EAAE,GAAG,IAAI,CAAC,CAAC;IAEhD,IAAI,KAAK,EAAE,OAAO,IAAI,KAAK,CAAC,OAAO,CAAC,SAAS,GAAG,EAAE,GAAG,OAAO,EAAE,CAAC;QAC7D,OAAO,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC;IAC7B,CAAC;IAED,MAAM,SAAS,GAAG,MAAM,SAAS,CAAC,OAAO,EAAE,KAAK,EAAE,EAAE,CAAC,CAAC;IACtD,MAAM,QAAQ,GAAG,OAAO,CAAC,SAAS,IAAI,qCAAqC,CAAC;IAE5E,MAAM,GAAG,GAAG,MAAM,SAAS,CAAC,QAAQ,EAAE;QACpC,MAAM,EAAE,MAAM;QACd,QAAQ,EAAE,OAAO;QACjB,OAAO,EAAE,EAAE,cAAc,EAAE,mCAAmC,EAAE;QAChE,IAAI,EAAE,IAAI,eAAe,CAAC;YACxB,UAAU,EAAE,6CAA6C;YACzD,SAAS;SACV,CAAC,CAAC,QAAQ,EAAE;KACd,CAAC,CAAC;IAEH,IAAI,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC;QACZ,MAAM,IAAI,KAAK,CAAC,iDAAiD,GAAG,CAAC,MAAM,MAAM,MAAM,GAAG,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC;IACvG,CAAC;IAED,MAAM,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC,IAAI,EAAE,CAAqD,CAAC;IACpF,IAAI,OAAO,IAAI,CAAC,YAAY,KAAK,QAAQ,IAAI,CAAC,IAAI,CAAC,YAAY,EAAE,CAAC;QAChE,MAAM,IAAI,KAAK,CAAC,0DAA0D,CAAC,CAAC;IAC9E,CAAC;IAED,MAAM,QAAQ,GAAG,OAAO,IAAI,CAAC,UAAU,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,CAAC;IAC9E,IAAI,KAAK;QAAE,KAAK,CAAC,OAAO,GAAG,EAAE,KAAK,EAAE,IAAI,CAAC,YAAY,EAAE,SAAS,EAAE,OAAO,GAAG,QAAQ,EAAE,CAAC;IAEvF,OAAO,IAAI,CAAC,YAAY,CAAC;AAC3B,CAAC"}

package/dist/glob.d.ts

@@ -0,0 +1,49 @@
+/**
+ * A small, bounded glob matcher.
+ *
+ * Trazum has zero runtime dependencies and that is a security property, not a
+ * preference: a prompt optimiser reads your prompts, and every dependency is
+ * someone else's code doing that too. So the config file's budget patterns are
+ * matched here rather than by pulling in a glob library.
+ *
+ * **It is deliberately not a regex translation.** Turning `**` into
+ * `(?:[^/]*\/)*` produces exactly the nested-quantifier shape that makes a
+ * pattern take exponential time on the wrong input, and these patterns come
+ * from a file in a repository — which on a pull request means from whoever
+ * opened it. A segment-wise dynamic program has no such shape: it is O(pattern
+ * segments x path segments), and each segment match is O(pattern chars x path
+ * chars) with no backtracking beyond a single remembered star.
+ *
+ * Supported, and nothing else:
+ *   `*`   any run of characters within one path segment
+ *   `**`  any number of whole path segments, including none
+ *   `?`   exactly one character within one path segment
+ */
+/**
+ * True when `path` matches `pattern`.
+ *
+ * Both sides are treated as relative paths with `/` separators; a leading `./`
+ * and duplicate separators are ignored, and backslashes are read as separators
+ * so a Windows-shaped path still matches a pattern written with slashes.
+ */
+export declare function matchGlob(pattern: string, path: string): boolean;
+/**
+ * How specific a pattern is, for deciding which of two matching patterns wins.
+ *
+ * "Most specific" needs a definition that is stated rather than felt, because
+ * a budget silently resolved from the wrong pattern is a budget nobody can
+ * debug. The rule: **more literal characters wins**, and a longer pattern
+ * breaks a tie. So `prompts/system.txt` beats `prompts/*.txt`, which beats
+ * `prompts/**`, which beats `**`.
+ */
+export declare function specificity(pattern: string): number;
+/**
+ * The most specific pattern in `patterns` that matches `path`, or null.
+ *
+ * Ties on specificity are broken by lexical order rather than by insertion
+ * order: object key order is easy to reorder by accident, and a budget that
+ * changes because two keys swapped places is the kind of bug that gets blamed
+ * on the tool.
+ */
+export declare function mostSpecificMatch(patterns: readonly string[], path: string): string | null;
+//# sourceMappingURL=glob.d.ts.map

package/dist/glob.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"glob.d.ts","sourceRoot":"","sources":["../src/glob.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;GAoBG;AAoDH;;;;;;GAMG;AACH,wBAAgB,SAAS,CAAC,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,GAAG,OAAO,CAmChE;AAED;;;;;;;;GAQG;AACH,wBAAgB,WAAW,CAAC,OAAO,EAAE,MAAM,GAAG,MAAM,CAOnD;AAED;;;;;;;GAOG;AACH,wBAAgB,iBAAiB,CAC/B,QAAQ,EAAE,SAAS,MAAM,EAAE,EAC3B,IAAI,EAAE,MAAM,GACX,MAAM,GAAG,IAAI,CAcf"}