opencodekit 0.17.13 → 0.18.0

package/dist/template/.opencode/tool/memory-search.ts

@@ -1,264 +0,0 @@
-import fs from "node:fs/promises";
-import path from "node:path";
-import { tool } from "@opencode-ai/plugin";
-import {
-	type ObservationType,
-	type SearchIndexResult,
-	checkFTS5Available,
-	searchObservationsFTS,
-} from "../plugin/lib/memory-db";
-
-// Fallback file-based search for non-SQLite content
-interface FileSearchResult {
-	file: string;
-	matches: { line: number; content: string }[];
-}
-
-async function searchDirectory(
-	dir: string,
-	pattern: RegExp,
-	results: FileSearchResult[],
-	baseDir: string,
-): Promise<void> {
-	try {
-		const entries = await fs.readdir(dir, { withFileTypes: true });
-
-		for (const entry of entries) {
-			const fullPath = path.join(dir, entry.name);
-
-			if (entry.isDirectory()) {
-				// Skip hidden directories and vector_db
-				if (entry.name.startsWith(".") || entry.name === "vector_db") {
-					continue;
-				}
-				await searchDirectory(fullPath, pattern, results, baseDir);
-			} else if (entry.name.endsWith(".md")) {
-				const content = await fs.readFile(fullPath, "utf-8");
-				const lines = content.split("\n");
-				const matches: { line: number; content: string }[] = [];
-
-				for (let index = 0; index < lines.length; index++) {
-					const line = lines[index];
-					if (pattern.test(line)) {
-						matches.push({
-							line: index + 1,
-							content: line.trim().substring(0, 150),
-						});
-					}
-				}
-
-				if (matches.length > 0) {
-					const relativePath = path.relative(baseDir, fullPath);
-					results.push({ file: relativePath, matches });
-				}
-			}
-		}
-	} catch {
-		// Directory doesn't exist or not readable
-	}
-}
-
-async function fallbackKeywordSearch(
-	query: string,
-	type: string | undefined,
-	limit: number,
-): Promise<FileSearchResult[]> {
-	const memoryDir = path.join(process.cwd(), ".opencode/memory");
-	const beadsDir = path.join(process.cwd(), ".beads/artifacts");
-	const globalMemoryDir = path.join(
-		process.env.HOME || "",
-		".config/opencode/memory",
-	);
-
-	// Create case-insensitive regex from query
-	let pattern: RegExp;
-	try {
-		pattern = new RegExp(query, "i");
-	} catch {
-		// Escape special chars if not valid regex
-		const escaped = query.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
-		pattern = new RegExp(escaped, "i");
-	}
-
-	const results: FileSearchResult[] = [];
-
-	// Handle type filtering
-	if (type === "beads") {
-		await searchDirectory(beadsDir, pattern, results, beadsDir);
-	} else if (type && type !== "all" && type !== "observations") {
-		const typeMap: Record<string, string> = {
-			handoffs: "handoffs",
-			research: "research",
-			templates: "_templates",
-		};
-		const subdir = typeMap[type];
-		if (subdir) {
-			const searchDir = path.join(memoryDir, subdir);
-			await searchDirectory(searchDir, pattern, results, memoryDir);
-		}
-	} else {
-		// Search all: memory + beads
-		await searchDirectory(memoryDir, pattern, results, memoryDir);
-		await searchDirectory(beadsDir, pattern, results, beadsDir);
-		await searchDirectory(globalMemoryDir, pattern, results, globalMemoryDir);
-	}
-
-	return results.slice(0, limit);
-}
-
-const TYPE_ICONS: Record<string, string> = {
-	decision: "🎯",
-	bugfix: "🐛",
-	feature: "✨",
-	pattern: "🔄",
-	discovery: "💡",
-	learning: "📚",
-	warning: "⚠️",
-};
-
-function formatCompactIndex(
-	results: SearchIndexResult[],
-	query: string,
-): string {
-	if (results.length === 0) {
-		return `No observations found for "${query}".\n\nTip: Use memory-search without query to see recent observations.`;
-	}
-
-	let output = `# Search Results: "${query}"\n\n`;
-	output += `Found **${results.length}** observation(s). Use \`memory-get\` for full details.\n\n`;
-	output += "| ID | Type | Title | Date |\n";
-	output += "|---|---|---|---|\n";
-
-	for (const result of results) {
-		const icon = TYPE_ICONS[result.type] || "📝";
-		const date = result.created_at.split("T")[0];
-		const title =
-			result.title.length > 50
-				? `${result.title.substring(0, 47)}...`
-				: result.title;
-		output += `| #${result.id} | ${icon} ${result.type} | ${title} | ${date} |\n`;
-	}
-
-	output += "\n## Snippets\n\n";
-	for (const result of results) {
-		const icon = TYPE_ICONS[result.type] || "📝";
-		output += `**#${result.id}** ${icon} ${result.title}\n`;
-		if (result.snippet) {
-			output += `> ${result.snippet}...\n`;
-		}
-		output += "\n";
-	}
-
-	output += "\n---\n";
-	output += `💡 **Next steps:**\n`;
-	output += `- \`memory-get({ ids: "${results
-		.map((r) => r.id)
-		.slice(0, 3)
-		.join(",")}" })\` - Get full details\n`;
-	output += `- \`memory-timeline({ anchor_id: ${results[0].id} })\` - See chronological context\n`;
-
-	return output;
-}
-
-function formatFallbackResults(
-	query: string,
-	results: FileSearchResult[],
-	limit: number,
-): string {
-	if (results.length === 0) {
-		return `No matches found for "${query}" in non-observation files.`;
-	}
-
-	let output = `# File Search: "${query}"\n\n`;
-	output += `Found ${results.length} file(s) with matches.\n\n`;
-
-	for (const result of results) {
-		output += `## ${result.file}\n\n`;
-		const matchesToShow = result.matches.slice(0, limit);
-		for (const match of matchesToShow) {
-			output += `- **Line ${match.line}:** ${match.content}\n`;
-		}
-		if (result.matches.length > limit) {
-			output += `- ... and ${result.matches.length - limit} more matches\n`;
-		}
-		output += "\n";
-	}
-
-	return output;
-}
-
-export default tool({
-	description: `Search memory across observations and markdown archives.
-	
-	Purpose:
-	- Fast, ranked search across all observations in SQLite (when FTS5 is available)
-	- Returns compact index (~50-100 tokens per result) for progressive disclosure
-	- Use memory-get for full details after identifying relevant observations
-	
-	FTS5 availability:
-	- Auto-detected at runtime; if unavailable, observation searches fall back to file scan
-	
-	Search modes and hints:
-	- "observations" (default): Best for decisions, bugs, learnings; uses FTS5 ranking when available
-	- "handoffs": Use for past session handoffs and summaries
-	- "research": Use for research notes and external findings
-	- "templates": Use for memory templates and boilerplate references
-	- "beads": Use for task artifacts in .beads/artifacts
-	- "all": Use when you are unsure where info lives; searches SQLite + markdown + beads
-	
-	Example:
-	memory-search({ query: "authentication" })
-	memory-search({ query: "auth", type: "decision", limit: 5 })`,
-	args: {
-		query: tool.schema.string().describe("Search query: keywords or phrase"),
-		type: tool.schema
-			.string()
-			.optional()
-			.describe(
-				"Filter by observation type (decision, bugfix, feature, pattern, discovery, learning, warning) or search scope (observations, handoffs, research, templates, beads, all)",
-			),
-		limit: tool.schema
-			.number()
-			.optional()
-			.describe("Max results (default: 10)"),
-	},
-	execute: async (args: { query: string; type?: string; limit?: number }) => {
-		const limit = args.limit || 10;
-
-		// Determine if we should use SQLite FTS5 or fallback
-		const observationTypes = [
-			"decision",
-			"bugfix",
-			"feature",
-			"pattern",
-			"discovery",
-			"learning",
-			"warning",
-		];
-		const isObservationType =
-			args.type && observationTypes.includes(args.type.toLowerCase());
-		const isObservationsScope =
-			!args.type || args.type === "observations" || isObservationType;
-
-		// Try SQLite FTS5 for observations
-		if (isObservationsScope && checkFTS5Available()) {
-			try {
-				const obsType = isObservationType
-					? (args.type?.toLowerCase() as ObservationType)
-					: undefined;
-				const results = searchObservationsFTS(args.query, {
-					type: obsType,
-					limit,
-				});
-				return formatCompactIndex(results, args.query);
-			} catch {
-				// FTS5 failed, fall through to file search
-				// Silently fall back to file-based search
-			}
-		}
-
-		// Fallback to file-based search for non-observation types or FTS5 failure
-		const results = await fallbackKeywordSearch(args.query, args.type, limit);
-		return formatFallbackResults(args.query, results, limit);
-	},
-});

package/dist/template/.opencode/tool/memory-timeline.ts

@@ -1,105 +0,0 @@
-import { tool } from "@opencode-ai/plugin";
-import {
-	type SearchIndexResult,
-	getTimelineAroundObservation,
-} from "../plugin/lib/memory-db";
-
-const TYPE_ICONS: Record<string, string> = {
-	decision: "🎯",
-	bugfix: "🐛",
-	feature: "✨",
-	pattern: "🔄",
-	discovery: "💡",
-	learning: "📚",
-	warning: "⚠️",
-};
-
-function formatTimelineResult(
-	anchor: {
-		id: number;
-		type: string;
-		title: string;
-		created_at: string;
-	} | null,
-	before: SearchIndexResult[],
-	after: SearchIndexResult[],
-): string {
-	if (!anchor) {
-		return "Anchor observation not found.";
-	}
-
-	let output = "# Timeline Context\n\n";
-
-	// Before (older observations)
-	if (before.length > 0) {
-		output += "## Earlier Observations\n\n";
-		for (const obs of before) {
-			const icon = TYPE_ICONS[obs.type] || "📝";
-			const date = obs.created_at.split("T")[0];
-			output += `- **#${obs.id}** ${icon} ${obs.title} _(${date})_\n`;
-			if (obs.snippet) {
-				output += `  ${obs.snippet.substring(0, 80)}...\n`;
-			}
-		}
-		output += "\n";
-	}
-
-	// Anchor
-	const anchorIcon = TYPE_ICONS[anchor.type] || "📝";
-	const anchorDate = anchor.created_at.split("T")[0];
-	output += `## ▶ Current: #${anchor.id}\n\n`;
-	output += `${anchorIcon} **${anchor.title}** _(${anchorDate})_\n\n`;
-
-	// After (newer observations)
-	if (after.length > 0) {
-		output += "## Later Observations\n\n";
-		for (const obs of after) {
-			const icon = TYPE_ICONS[obs.type] || "📝";
-			const date = obs.created_at.split("T")[0];
-			output += `- **#${obs.id}** ${icon} ${obs.title} _(${date})_\n`;
-			if (obs.snippet) {
-				output += `  ${obs.snippet.substring(0, 80)}...\n`;
-			}
-		}
-	}
-
-	return output;
-}
-
-export default tool({
-	description: `Get chronological context around an observation.
-	
-	Purpose:
-	- Progressive disclosure: see what was happening before/after a specific observation
-	- Understand decision context over time
-	- Navigate memory timeline
-	
-	Example:
-	memory-timeline({ anchor_id: 42, depth_before: 5, depth_after: 5 })`,
-	args: {
-		anchor_id: tool.schema
-			.number()
-			.describe("ID of the observation to get context around"),
-		depth_before: tool.schema
-			.number()
-			.optional()
-			.describe("Number of earlier observations to include (default: 5)"),
-		depth_after: tool.schema
-			.number()
-			.optional()
-			.describe("Number of later observations to include (default: 5)"),
-	},
-	execute: async (args: {
-		anchor_id: number;
-		depth_before?: number;
-		depth_after?: number;
-	}) => {
-		const { anchor, before, after } = getTimelineAroundObservation(
-			args.anchor_id,
-			args.depth_before ?? 5,
-			args.depth_after ?? 5,
-		);
-
-		return formatTimelineResult(anchor, before, after);
-	},
-});

package/dist/template/.opencode/tool/memory-update.ts

@@ -1,63 +0,0 @@
-import { tool } from "@opencode-ai/plugin";
-import { upsertMemoryFile } from "../plugin/lib/memory-db.js";
-
-export default tool({
-	description: `Update memory files with new learnings, progress, or context.
-	
-	Purpose:
-	- Write or append to project memory in SQLite
-	- Supports subdirectories (e.g., 'research/2024-01-topic')
-	- Two modes: 'replace' (overwrite) or 'append' (add to end)
-	
-	Example:
-	memory-update({ file: "research/session-findings", content: "..." })
-	memory-update({ file: "handoffs/phase-2", content: "...", mode: "append" })`,
-	args: {
-		file: tool.schema
-			.string()
-			.describe(
-				"Memory file to update: handoffs/YYYY-MM-DD-phase, research/YYYY-MM-DD-topic",
-			),
-		content: tool.schema
-			.string()
-			.describe("Content to write or append to the memory file"),
-		mode: tool.schema
-			.string()
-			.optional()
-			.default("replace")
-			.describe(
-				"Update mode: 'replace' (overwrite file) or 'append' (add to end).",
-			),
-	},
-	execute: async (args: { file: string; content: string; mode?: string }) => {
-		// Normalize file path: strip existing .md extension
-		const normalizedFile = args.file.replace(/\.md$/i, "");
-		const mode = (args.mode || "replace") as "replace" | "append";
-
-		const timestamp = new Date().toISOString();
-		let finalContent: string;
-
-		if (mode === "append") {
-			finalContent = `\n\n---\n**Updated:** ${timestamp}\n\n${args.content}`;
-		} else {
-			// Replace mode - update timestamp placeholder if present
-			finalContent = args.content.replace(
-				/\*\*Last Updated:\*\* \[Timestamp\]/,
-				`**Last Updated:** ${timestamp}`,
-			);
-		}
-
-		try {
-			// Store in SQLite (single source of truth)
-			upsertMemoryFile(normalizedFile, finalContent, mode);
-
-			const action = mode === "append" ? "Appended to" : "Updated";
-			return `✓ ${action} ${normalizedFile}`;
-		} catch (error) {
-			if (error instanceof Error) {
-				return `Error updating memory: ${error.message}`;
-			}
-			return "Unknown error updating memory file";
-		}
-	},
-});