one-code-extension 0.1.4 → 0.3.1

package/extensions/lib/skill-overrides.ts

@@ -0,0 +1,86 @@
+/**
+ * Per-skill availability overrides (pure fs, no pi imports).
+ *
+ * `<plugin root>/skill-overrides.json`: `{"<scope>:<skill name>": SkillState}`.
+ * Mirrors Claude Code's `skillOverrides` setting — four states controlling how
+ * much of a skill reaches the model and whether the user can still run it:
+ *
+ *   on         name + description listed to the model; user-invocable
+ *   name-only  name listed (no description, saving context tokens); user-invocable
+ *   user-only  hidden from the model (won't auto-trigger); user-invocable via /name
+ *   off        hidden from the model AND refused on explicit invocation
+ *
+ * Absent key = "on" (every discovered skill is fully available by default). The
+ * /skills panel cycles these states for project/user skills; plugin skills are
+ * "locked" (their on/off is managed via /plugins, stored here under the
+ * "plugin" scope as on/off only). Legacy boolean values round-trip as on/off.
+ *
+ * The boolean-facing `isSkillEnabled`/`setSkillOverride` wrappers keep the
+ * /plugins panel and the plugin-skill filter (which only ever enable/disable)
+ * working against the same store.
+ */
+
+import { join } from "node:path";
+import { readJsonFile, writeJsonAtomic } from "./atomic-write.ts";
+
+const FILE = "skill-overrides.json";
+
+export type SkillScope = "user" | "project" | "plugin";
+export type SkillState = "on" | "name-only" | "user-only" | "off";
+
+/** Cycle order for the /skills panel (space/enter advances). */
+export const SKILL_STATES: readonly SkillState[] = ["on", "name-only", "user-only", "off"];
+
+export function skillOverrideKey(scope: SkillScope, name: string): string {
+	return `${scope}:${name}`;
+}
+
+/** Read the store as states, coercing legacy booleans (true → on, false → off). */
+export function readSkillStates(root: string): Record<string, SkillState> {
+	const parsed = readJsonFile<Record<string, unknown>>(join(root, FILE));
+	if (!parsed || typeof parsed !== "object") return {};
+	const map: Record<string, SkillState> = {};
+	for (const [key, value] of Object.entries(parsed)) {
+		if (typeof value === "string" && (SKILL_STATES as readonly string[]).includes(value)) {
+			map[key] = value as SkillState;
+		} else if (value === true) {
+			map[key] = "on";
+		} else if (value === false) {
+			map[key] = "off";
+		}
+	}
+	return map;
+}
+
+export function setSkillState(root: string, key: string, state: SkillState): void {
+	const existing = readJsonFile<Record<string, unknown>>(join(root, FILE)) ?? {};
+	writeJsonAtomic(join(root, FILE), { ...existing, [key]: state });
+}
+
+export function skillStateFor(states: Record<string, SkillState>, key: string): SkillState {
+	return states[key] ?? "on";
+}
+
+/** Advance to the next state in the cycle (wraps). */
+export function nextSkillState(state: SkillState): SkillState {
+	return SKILL_STATES[(SKILL_STATES.indexOf(state) + 1) % SKILL_STATES.length];
+}
+
+/** How much of the skill the model's listing should carry for a given state. */
+export function skillListingVisibility(state: SkillState): "full" | "name" | "hidden" {
+	return state === "on" ? "full" : state === "name-only" ? "name" : "hidden";
+}
+
+// --- Boolean-facing compatibility (plugins panel + plugin-skill filter) ---
+
+/** Enabled = anything but "off"; the plugin filter and Installed tab use this. */
+export function isSkillEnabled(states: Record<string, SkillState>, key: string): boolean {
+	return skillStateFor(states, key) !== "off";
+}
+
+export function setSkillOverride(root: string, key: string, enabled: boolean): void {
+	setSkillState(root, key, enabled ? "on" : "off");
+}
+
+
+

package/extensions/lib/skill-scan.ts

@@ -0,0 +1,110 @@
+/**
+ * Skill discovery for the pre-first-turn /skills panel and the bare `/<skill>`
+ * command registration (pure fs). (The /plugins Installed tab builds its rows
+ * from discoverPlugins directly and no longer calls this.)
+ *
+ * Scans the SAME directories claude-compat feeds to pi, in pi's order —
+ * `~/.claude/skills`, `~/.agents/skills`, `<cwd>/.claude/skills`,
+ * `<cwd>/.agents/skills`, `<agentDir>/skills`, then the bundled catalog last
+ * (so it loses a name collision, as it does in pi) — and merges the plugin
+ * skills the caller resolved. Rows carry the SKILL.md path (for the ~token
+ * size estimate) and the scope used by skill-overrides keys. Before the first
+ * turn this scan is the only skill list One Code has (pi resolves skills per
+ * turn), so a root missing here is a skill with no `/name` command at startup.
+ */
+
+import { existsSync, readdirSync, statSync } from "node:fs";
+import { dirname, join } from "node:path";
+import { fileURLToPath } from "node:url";
+import type { PluginSkill } from "./plugins.ts";
+import type { SkillScope } from "./skill-overrides.ts";
+import { claudeUserDir } from "./paths.ts";
+
+/** The skill catalog shipped in this package: `<package>/skills` (Claude Code's self-contained built-in skills). */
+export const BUNDLED_SKILLS_DIR = join(dirname(fileURLToPath(import.meta.url)), "..", "..", "skills");
+
+export interface ScannedSkill {
+	name: string;
+	path: string;
+	scope: SkillScope;
+}
+
+/**
+ * The one scope-classification rule for a skill path — user skills live under
+ * `~/.claude/skills` AND `<agentDir>/skills`; everything else is project
+ * scope. Both the scanner and the skill extension classify through this, so a
+ * skill never shows different scopes in /plugins vs the skill tool.
+ */
+export function scopeForPath(path: string, home: string, agentDir: string): SkillScope {
+	if (path.startsWith(join(claudeUserDir(home), "skills")) || path.startsWith(join(agentDir, "skills"))) return "user";
+	return "project";
+}
+
+function scanDir(dir: string, scope: SkillScope, into: Map<string, ScannedSkill>): void {
+	if (!existsSync(dir)) return;
+	let entries: string[];
+	try {
+		entries = readdirSync(dir);
+	} catch {
+		return;
+	}
+	for (const entry of entries) {
+		const skillFile = join(dir, entry, "SKILL.md");
+		// existsSync follows symlinked skill directories, which isDirectory() would miss.
+		if (existsSync(skillFile) && !into.has(`${scope}:${entry}`)) {
+			into.set(`${scope}:${entry}`, { name: entry, path: skillFile, scope });
+		}
+	}
+}
+
+export function scanSkills(
+	cwd: string,
+	home: string,
+	agentDir: string,
+	pluginSkills: PluginSkill[],
+	bundledSkillsDir?: string,
+): ScannedSkill[] {
+	const skills = new Map<string, ScannedSkill>();
+	scanDir(join(claudeUserDir(home), "skills"), "user", skills);
+	scanDir(join(home, ".agents", "skills"), "user", skills);
+	scanDir(join(cwd, ".claude", "skills"), "project", skills);
+	scanDir(join(cwd, ".agents", "skills"), "project", skills);
+	scanDir(join(agentDir, "skills"), "user", skills);
+	if (bundledSkillsDir) scanDir(bundledSkillsDir, scopeForPath(bundledSkillsDir, home, agentDir), skills);
+	for (const skill of pluginSkills) {
+		skills.set(`plugin:${skill.name}`, { name: skill.name, path: skill.path, scope: "plugin" });
+	}
+	return [...skills.values()].sort((a, b) => a.name.localeCompare(b.name));
+}
+
+/**
+ * Names of the `.claude/commands` prompt templates (`<name>.md`) pi will expose
+ * as `/<name>` — the same dirs claude-compat feeds pi (`~/.claude/commands`,
+ * `<cwd>/.claude/commands`) plus pi's own `<agentDir>/prompts`. pi resolves
+ * templates per turn, after `session_start`, so a bare skill command registered
+ * then would silently shadow a same-named template for the whole session;
+ * this pre-scan lets the registration skip those names.
+ */
+export function promptTemplateNames(cwd: string, home: string, agentDir: string): string[] {
+	const names = new Set<string>();
+	for (const dir of [join(claudeUserDir(home), "commands"), join(cwd, ".claude", "commands"), join(agentDir, "prompts")]) {
+		if (!existsSync(dir)) continue;
+		let entries: string[];
+		try {
+			entries = readdirSync(dir);
+		} catch {
+			continue;
+		}
+		for (const entry of entries) if (entry.endsWith(".md")) names.add(entry.slice(0, -3));
+	}
+	return [...names];
+}
+
+/** Rough token estimate from the SKILL.md byte size (~4 bytes/token). */
+export function estimateSkillTokens(path: string): number {
+	try {
+		return Math.max(1, Math.round(statSync(path).size / 4));
+	} catch {
+		return 0;
+	}
+}

package/extensions/lib/text-width.ts

@@ -0,0 +1,267 @@
+/**
+ * Terminal-column width for text, matching pi-tui's own measurement.
+ *
+ * pi-tui measures a rendered line in terminal *columns* — a CJK ideograph or
+ * an emoji is two columns, a combining mark zero — and in regular TUI mode it
+ * *crashes the whole app* on a line wider than the terminal ("Rendered line
+ * exceeds terminal width"). Our width helpers historically counted code points
+ * (`[...text].length`) or UTF-16 units (`line.length`), so a line of N wide
+ * glyphs at N ≤ width was "fitted" at 2N columns and killed pi (TUI-REVIEW H1).
+ *
+ * This module reproduces pi-tui's `visibleWidth`/`graphemeWidth`
+ * (`@earendil-works/pi-tui` `dist/utils.js`) so every helper on top of it
+ * truncates by the same measure pi-tui validates against. It stays a pure,
+ * zero-dependency module: the East Asian wide/fullwidth ranges below are the
+ * `wideRanges`/`fullwidthRanges` tables from `get-east-asian-width` (MIT,
+ * © Sindre Sorhus — pi-tui's own width dependency), embedded verbatim so the
+ * 2-vs-1 decision is byte-identical to pi's without importing pi or adding a
+ * package. Grapheme segmentation and the mark/emoji classification use V8's
+ * built-in `Intl.Segmenter` and Unicode property regexes.
+ */
+
+// --- East Asian wide/fullwidth lookup (get-east-asian-width v1.6.0, Unicode 16) ---
+// Flat, sorted [start, end] inclusive pairs. `eastAsianWidth(cp)` returns 2 when
+// cp is fullwidth or wide, else 1 — the only distinction pi-tui's grapheme
+// width needs (ambiguous is treated as narrow, `ambiguousAsWide` off).
+const FULLWIDTH_MIN = 12288;
+const FULLWIDTH_MAX = 65510;
+// prettier-ignore
+const FULLWIDTH_RANGES = [12288, 12288, 65281, 65376, 65504, 65510];
+const WIDE_MIN = 4352;
+const WIDE_MAX = 262141;
+// prettier-ignore
+const WIDE_RANGES = [
+	4352, 4447, 8986, 8987, 9001, 9002, 9193, 9196, 9200, 9200, 9203, 9203, 9725, 9726, 9748, 9749, 9776, 9783,
+	9800, 9811, 9855, 9855, 9866, 9871, 9875, 9875, 9889, 9889, 9898, 9899, 9917, 9918, 9924, 9925, 9934, 9934,
+	9940, 9940, 9962, 9962, 9970, 9971, 9973, 9973, 9978, 9978, 9981, 9981, 9989, 9989, 9994, 9995, 10024,
+	10024, 10060, 10060, 10062, 10062, 10067, 10069, 10071, 10071, 10133, 10135, 10160, 10160, 10175, 10175,
+	11035, 11036, 11088, 11088, 11093, 11093, 11904, 11929, 11931, 12019, 12032, 12245, 12272, 12287, 12289,
+	12350, 12353, 12438, 12441, 12543, 12549, 12591, 12593, 12686, 12688, 12773, 12783, 12830, 12832, 12871,
+	12880, 42124, 42128, 42182, 43360, 43388, 44032, 55203, 63744, 64255, 65040, 65049, 65072, 65106, 65108,
+	65126, 65128, 65131, 94176, 94180, 94192, 94198, 94208, 101589, 101631, 101662, 101760, 101874, 110576,
+	110579, 110581, 110587, 110589, 110590, 110592, 110882, 110898, 110898, 110928, 110930, 110933, 110933,
+	110948, 110951, 110960, 111355, 119552, 119638, 119648, 119670, 126980, 126980, 127183, 127183, 127374,
+	127374, 127377, 127386, 127488, 127490, 127504, 127547, 127552, 127560, 127568, 127569, 127584, 127589,
+	127744, 127776, 127789, 127797, 127799, 127868, 127870, 127891, 127904, 127946, 127951, 127955, 127968,
+	127984, 127988, 127988, 127992, 128062, 128064, 128064, 128066, 128252, 128255, 128317, 128331, 128334,
+	128336, 128359, 128378, 128378, 128405, 128406, 128420, 128420, 128507, 128591, 128640, 128709, 128716,
+	128716, 128720, 128722, 128725, 128728, 128732, 128735, 128747, 128748, 128756, 128764, 128992, 129003,
+	129008, 129008, 129292, 129338, 129340, 129349, 129351, 129535, 129648, 129660, 129664, 129674, 129678,
+	129734, 129736, 129736, 129741, 129756, 129759, 129770, 129775, 129784, 131072, 196605, 196608, 262141,
+];
+
+/** Binary search over a flat, sorted array of inclusive [start, end] pairs. */
+function inRange(ranges: number[], codePoint: number): boolean {
+	let low = 0;
+	let high = ranges.length / 2 - 1;
+	while (low <= high) {
+		const mid = (low + high) >> 1;
+		const i = mid * 2;
+		if (codePoint < ranges[i]) high = mid - 1;
+		else if (codePoint > ranges[i + 1]) low = mid + 1;
+		else return true;
+	}
+	return false;
+}
+
+/** 2 for East Asian fullwidth/wide code points, 1 otherwise. */
+function eastAsianWidth(codePoint: number): number {
+	if (codePoint >= FULLWIDTH_MIN && codePoint <= FULLWIDTH_MAX && inRange(FULLWIDTH_RANGES, codePoint)) return 2;
+	if (codePoint >= WIDE_MIN && codePoint <= WIDE_MAX && inRange(WIDE_RANGES, codePoint)) return 2;
+	return 1;
+}
+
+// --- Grapheme classification (mirrors pi-tui dist/utils.js) ---
+// `Intl.Segmenter` is a V8 built-in; the ES2023 lib does not always type it.
+const Segmenter: new (locale?: string, options?: { granularity: string }) => {
+	segment(input: string): Iterable<{ segment: string }>;
+} = (Intl as unknown as { Segmenter: any }).Segmenter;
+const graphemeSegmenter = new Segmenter(undefined, { granularity: "grapheme" });
+
+// Built via `new RegExp(…, "v")` so the `v` (Unicode sets) flag — needed for
+// `\p{RGI_Emoji}` and the set-subtraction below, and supported at runtime on
+// Node 22+/26 — is not rejected by the ES2023 type target's static flag check.
+const zeroWidthRegex = new RegExp(String.raw`^(?:\p{Default_Ignorable_Code_Point}|\p{Control}|\p{Mark}|\p{Surrogate})+$`, "v");
+const leadingNonPrintingRegex = new RegExp(String.raw`^[\p{Default_Ignorable_Code_Point}\p{Control}\p{Format}\p{Mark}\p{Surrogate}]+`, "v");
+const nonPrintingCharRegex = new RegExp(String.raw`^(?:\p{Default_Ignorable_Code_Point}|\p{Control}|\p{Format}|\p{Mark}|\p{Surrogate})$`, "v");
+const markCharRegex = new RegExp(String.raw`^\p{Mark}$`, "v");
+// Marks that terminals allocate cells for when attached to a base character.
+const terminalSpacingMarkRegex = new RegExp(
+	String.raw`^(?:[\p{Spacing_Mark}--[᜴〮〯]]|[ٟཿါာေဳ-ဵး်-ှ])+$`,
+	"v",
+);
+const rgiEmojiRegex = new RegExp(String.raw`^\p{RGI_Emoji}$`, "v");
+
+/** Fast heuristic to skip the expensive RGI_Emoji test for clearly-non-emoji clusters. */
+function couldBeEmoji(segment: string): boolean {
+	const cp = segment.codePointAt(0) ?? 0;
+	return (
+		(cp >= 0x1f000 && cp <= 0x1fbff) ||
+		(cp >= 0x2300 && cp <= 0x23ff) ||
+		(cp >= 0x2600 && cp <= 0x27bf) ||
+		(cp >= 0x2b50 && cp <= 0x2b55) ||
+		segment.includes("️") ||
+		segment.length > 2
+	);
+}
+
+/** Terminal columns occupied by one grapheme cluster. Mirrors pi-tui's graphemeWidth. */
+export function graphemeWidth(segment: string): number {
+	if (segment === "\t") return 3;
+	if (terminalSpacingMarkRegex.test(segment)) return [...segment].length;
+	if (zeroWidthRegex.test(segment)) return 0;
+	if (couldBeEmoji(segment) && rgiEmojiRegex.test(segment)) return 2;
+	const base = segment.replace(leadingNonPrintingRegex, "");
+	const cp = base.codePointAt(0);
+	if (cp === undefined) return 0;
+	// Regional-indicator symbols render as full-width emoji even in isolation.
+	if (cp >= 0x1f1e6 && cp <= 0x1f1ff) return 2;
+	let width = eastAsianWidth(cp);
+	let followsMark = false;
+	for (const char of [...base].slice(1)) {
+		if (terminalSpacingMarkRegex.test(char)) {
+			width += 1;
+			followsMark = false;
+		} else if (markCharRegex.test(char)) {
+			followsMark = true;
+		} else if (!nonPrintingCharRegex.test(char)) {
+			const c = char.codePointAt(0)!;
+			if (followsMark || (c >= 0xff00 && c <= 0xffef)) width += eastAsianWidth(c);
+			followsMark = false;
+		}
+	}
+	return width;
+}
+
+// --- ANSI / OSC / APC escape handling ---
+// A single terminal control sequence starting at index 0 of `s`, or null.
+// Covers CSI (styling/cursor), OSC (hyperlinks, prompt markers, terminated by
+// BEL or ST) and APC (terminated by ST) — the same families pi-tui strips.
+function escapeAt(s: string, i: number): number {
+	if (s.charCodeAt(i) !== 0x1b) return 0;
+	const next = s[i + 1];
+	if (next === "[") {
+		// CSI: ESC [ ... final byte 0x40–0x7E
+		let j = i + 2;
+		while (j < s.length) {
+			const c = s.charCodeAt(j);
+			j++;
+			if (c >= 0x40 && c <= 0x7e) return j - i;
+		}
+		return s.length - i;
+	}
+	if (next === "]" || next === "_" || next === "P" || next === "^") {
+		// OSC / APC / DCS / PM: ESC <intro> ... terminated by BEL or ST (ESC \)
+		let j = i + 2;
+		while (j < s.length) {
+			if (s.charCodeAt(j) === 0x07) return j + 1 - i; // BEL
+			if (s.charCodeAt(j) === 0x1b && s[j + 1] === "\\") return j + 2 - i; // ST
+			j++;
+		}
+		return s.length - i;
+	}
+	// Other two-byte escapes (ESC c, ESC =, …)
+	return next ? 2 : 1;
+}
+
+const PRINTABLE_ASCII = /^[\x20-\x7e]*$/;
+
+/**
+ * Visible width of a string in terminal columns: ANSI/OSC/APC escapes count
+ * zero, tabs count three, wide glyphs and emoji count two, combining marks
+ * zero. This is the measure pi-tui validates a rendered line against.
+ */
+export function visibleWidth(text: string): number {
+	if (text.length === 0) return 0;
+	if (PRINTABLE_ASCII.test(text)) return text.length;
+	let clean = "";
+	for (let i = 0; i < text.length; ) {
+		const esc = escapeAt(text, i);
+		if (esc) {
+			i += esc;
+			continue;
+		}
+		clean += text[i];
+		i++;
+	}
+	// A painted line is usually plain-ASCII text wrapped in colour escapes; once
+	// the escapes are gone the fast path applies again, skipping the segmenter and
+	// per-grapheme regexes on the common case (every coloured menu/status line).
+	if (PRINTABLE_ASCII.test(clean)) return clean.length;
+	let width = 0;
+	for (const { segment } of graphemeSegmenter.segment(clean)) width += graphemeWidth(segment);
+	return width;
+}
+
+/**
+ * Longest prefix of PLAIN (escape-free) text whose visible width is ≤ `maxCols`,
+ * with the width actually taken. Splits on grapheme boundaries, so a wide glyph
+ * is kept whole or dropped whole — never half-included past the limit. Plain text
+ * is the escape-free case of `fitPainted`, so this delegates rather than keeping
+ * a second copy of the grapheme walk.
+ */
+export function sliceColumns(text: string, maxCols: number): { text: string; width: number } {
+	return fitPainted(text, maxCols);
+}
+
+/**
+ * Hard-wrap text into chunks each ≤ `columns` visible columns, on grapheme
+ * boundaries. A single glyph wider than the whole budget occupies its own chunk
+ * (nothing can be dropped without losing content). The shared engine behind
+ * `wrapPlainText` and the subagent panel's prose wrap.
+ */
+export function hardWrapColumns(text: string, columns: number): string[] {
+	const cols = Math.max(1, columns);
+	if (text.length === 0) return [""];
+	const out: string[] = [];
+	let rest = text;
+	// Each step consumes a fitPainted prefix and advances; `fitPainted` reports
+	// whether it took the whole remainder, so no separate full-width rescan is
+	// needed (that rescan made wrapping O(n²) — TUI-REVIEW /simplify).
+	for (;;) {
+		const { text: head } = fitPainted(rest, cols);
+		const chunk = head || [...rest][0] || rest; // guarantee progress on an over-wide glyph
+		if (chunk.length >= rest.length) {
+			out.push(rest);
+			return out;
+		}
+		out.push(chunk);
+		rest = rest.slice(chunk.length);
+	}
+}
+
+/**
+ * Longest prefix of a PAINTED line (interspersed ANSI/OSC escapes) whose visible
+ * width is ≤ `maxCols`. Escapes are kept verbatim and cost zero columns; visible
+ * runs are cut on grapheme boundaries. Returns the prefix and its visible width;
+ * the caller supplies its own reset/ellipsis.
+ */
+export function fitPainted(line: string, maxCols: number): { text: string; width: number } {
+	if (maxCols <= 0) return { text: "", width: 0 };
+	if (PRINTABLE_ASCII.test(line)) {
+		const clipped = line.slice(0, maxCols);
+		return { text: clipped, width: clipped.length };
+	}
+	let out = "";
+	let width = 0;
+	let i = 0;
+	while (i < line.length) {
+		const esc = escapeAt(line, i);
+		if (esc) {
+			out += line.slice(i, i + esc);
+			i += esc;
+			continue;
+		}
+		// Consume the visible run up to the next escape, segment it into graphemes.
+		let end = i;
+		while (end < line.length && line.charCodeAt(end) !== 0x1b) end++;
+		for (const { segment } of graphemeSegmenter.segment(line.slice(i, end))) {
+			const w = graphemeWidth(segment);
+			if (width + w > maxCols) return { text: out, width };
+			out += segment;
+			width += w;
+		}
+		i = end;
+	}
+	return { text: out, width };
+}