@abranjith/spec-lite 0.0.1 → 0.0.3

package/README.md

@@ -14,7 +14,7 @@ spec-lite is a set of **modular prompt files** — each defining a specialist AI
 - **Modular** — Use one sub-agent or all of them. Skip what you don't need.
 - **Unopinionated** — Adapts to any project type (web, CLI, library, desktop, pipeline), any language, any stack.
 - **Finite-scoped** — Each sub-agent has one job, clear inputs, and a concrete output artifact.
-- **Memory-first** — Cross-cutting standards (coding conventions, architecture, testing, security) live in `.spec/memory.md` — the single source of truth read by every sub-agent.
+- **Memory-first** — Cross-cutting standards (coding conventions, architecture, testing, security) live in `.spec-lite/memory.md` — the single source of truth read by every sub-agent.
 - **Provider-agnostic** — Works with GitHub Copilot, Claude Code, or any LLM via generic mode.
 
 ---
@@ -22,7 +22,7 @@ spec-lite is a set of **modular prompt files** — each defining a specialist AI
 ## Installation
 
 ```bash
-npm install -g spec-lite
+npm install -g @abranjith/spec-lite
 ```
 
 Requires Node.js 18+.
@@ -47,7 +47,7 @@ The CLI will walk you through a short **project profile questionnaire** (languag
 1. Write agent prompt files to the correct location for your AI tool
 2. Inject your tech-stack context into every prompt's `<!-- project-context -->` block
 3. Copy a curated **best-practice snippet** for your stack into `.spec-lite/stacks/`
-4. Create the `.spec/` directory structure for agent outputs
+4. Create the `.spec-lite/` directory structure for agent outputs
 5. Save a `.spec-lite.json` config (including your project profile) to track your setup
 
 After init completes, run **`/memorize bootstrap`** (see below) to let the LLM auto-generate a comprehensive `memory.md` from your codebase.
@@ -88,12 +88,12 @@ More providers (Cursor, Windsurf, Cline, Zed) coming soon.
 
 ## Memory-First Architecture
 
-spec-lite uses a **memory-first** approach: cross-cutting concerns that every sub-agent needs — coding standards, architecture patterns, testing conventions, security guidelines, logging strategy — live in a single file: **`.spec/memory.md`**.
+spec-lite uses a **memory-first** approach: cross-cutting concerns that every sub-agent needs — coding standards, architecture patterns, testing conventions, security guidelines, logging strategy — live in a single file: **`.spec-lite/memory.md`**.
 
 | Source | Purpose | Authority |
 |--------|---------|-----------|
-| `.spec/memory.md` | Cross-cutting standards & conventions | **Primary** — authoritative for all sub-agents |
-| `.spec/plan.md` or `.spec/plan_<name>.md` | Project-specific blueprint(s) & task breakdown | Overrides memory only with explicit justification |
+| `.spec-lite/memory.md` | Cross-cutting standards & conventions | **Primary** — authoritative for all sub-agents |
+| `.spec-lite/plan.md` or `.spec-lite/plan_<name>.md` | Project-specific blueprint(s) & task breakdown | Overrides memory only with explicit justification |
 | User instruction | Ad-hoc guidance in chat | Highest priority (trumps both) |
 
 ### Bootstrap Flow
@@ -121,45 +121,47 @@ spec_help (anytime)
 
                   ┌─ /memorize bootstrap (one-time setup)
                   ▼
-Brainstorm ─→ Planner ─→ Feature (×N) ─→ Reviews ─→ Tests ─→ DevOps ─→ Docs
+Brainstorm ─→ Planner ─→ Architect ─→ Feature (×N) ─→ Reviews ─→ Tests ─→ DevOps ─→ Docs
                 │                          ├─ Code Review
                 │                          ├─ Security Audit
                 ▼                          └─ Performance Review
             TODO.md (living backlog)
 ```
 
-All sub-agents read `.spec/memory.md` first for standing instructions, then the relevant plan (`.spec/plan.md` or `.spec/plan_<name>.md`) for project-specific context. Complex projects can have multiple named plans — one per domain (e.g., `plan_order_management.md`, `plan_catalog.md`). Not every project needs every sub-agent. Start with the Planner if you already have requirements. Use `spec-lite list` or the spec_help sub-agent to understand the pipeline.
+All sub-agents read `.spec-lite/memory.md` first for standing instructions, then the relevant plan (`.spec-lite/plan.md` or `.spec-lite/plan_<name>.md`) for project-specific context. Complex projects can have multiple named plans — one per domain (e.g., `plan_order_management.md`, `plan_catalog.md`). Not every project needs every sub-agent. Start with the Planner if you already have requirements. Use `spec-lite list` or the spec_help sub-agent to understand the pipeline.
 
 ## Sub-Agent Prompt Files
 
 | File | Sub-Agent | What It Does | Output |
 |------|-----------|-------------|--------|
 | [spec_help.md](prompts/spec_help.md) | Spec Help | Navigator — explains which sub-agent to use and when | Interactive guidance |
-| [brainstorm.md](prompts/brainstorm.md) | Brainstorm | Back-and-forth ideation partner that refines vague ideas | `.spec/brainstorm.md` |
-| [planner.md](prompts/planner.md) | Planner | Creates a detailed technical blueprint (living document) | `.spec/plan.md` or `.spec/plan_<name>.md` |
-| [feature.md](prompts/feature.md) | Feature | 3-phase lifecycle: explore → tasks → implement+test+docs | `.spec/features/feature_<name>.md` |
-| [code_review.md](prompts/code_review.md) | Code Review | Reviews code for correctness, architecture, readability | `.spec/reviews/code_review_<name>.md` |
-| [security_audit.md](prompts/security_audit.md) | Security Audit | Threat-models and scans for vulnerabilities | `.spec/reviews/security_audit.md` |
-| [performance_review.md](prompts/performance_review.md) | Performance Review | Identifies bottlenecks and optimization opportunities | `.spec/reviews/performance_review.md` |
-| [integration_tests.md](prompts/integration_tests.md) | Integration Tests | Writes traceable integration test scenarios from feature specs | `.spec/features/integration_tests_<name>.md` |
-| [unit_tests.md](prompts/unit_tests.md) | Unit Tests | Generates comprehensive unit tests with edge-case coverage and smart coverage exclusions | `.spec/features/unit_tests_<name>.md` |
-| [devops.md](prompts/devops.md) | DevOps | Sets up Docker, CI/CD, environments, and deployment | `.spec/devops/` + infra files |
-| [fix.md](prompts/fix.md) | Fix | Debugs issues with root cause analysis + regression tests | `.spec/reviews/fix_<issue>.md` |
+| [brainstorm.md](prompts/brainstorm.md) | Brainstorm | Back-and-forth ideation partner that refines vague ideas | `.spec-lite/brainstorm.md` |
+| [planner.md](prompts/planner.md) | Planner | Creates a detailed technical blueprint (living document) | `.spec-lite/plan.md` or `.spec-lite/plan_<name>.md` |
+| [architect.md](prompts/architect.md) | Architect | Designs cloud infrastructure, database strategy, and scaling architecture | `.spec-lite/architect_<name>.md` |
+| [feature.md](prompts/feature.md) | Feature | 3-phase lifecycle: explore → tasks → implement+test+docs | `.spec-lite/features/feature_<name>.md` |
+| [code_review.md](prompts/code_review.md) | Code Review | Reviews code for correctness, architecture, readability | `.spec-lite/reviews/code_review_<name>.md` |
+| [security_audit.md](prompts/security_audit.md) | Security Audit | Threat-models and scans for vulnerabilities | `.spec-lite/reviews/security_audit.md` |
+| [performance_review.md](prompts/performance_review.md) | Performance Review | Identifies bottlenecks and optimization opportunities | `.spec-lite/reviews/performance_review.md` |
+| [integration_tests.md](prompts/integration_tests.md) | Integration Tests | Writes traceable integration test scenarios from feature specs | `.spec-lite/features/integration_tests_<name>.md` |
+| [unit_tests.md](prompts/unit_tests.md) | Unit Tests | Generates comprehensive unit tests with edge-case coverage and smart coverage exclusions | `.spec-lite/features/unit_tests_<name>.md` |
+| [devops.md](prompts/devops.md) | DevOps | Sets up Docker, CI/CD, environments, and deployment | `.spec-lite/devops/` + infra files |
+| [fix.md](prompts/fix.md) | Fix | Debugs issues with root cause analysis + regression tests | `.spec-lite/reviews/fix_<issue>.md` |
 | [technical_docs.md](prompts/technical_docs.md) | Technical Docs | Creates architecture docs, API references, setup guides | Technical documentation |
 | [readme.md](prompts/readme.md) | README | Writes the project README | `README.md` |
-| [memorize.md](prompts/memorize.md) | Memorize | Manages `.spec/memory.md` — standing instructions for all agents. Use `/memorize bootstrap` to auto-generate. | `.spec/memory.md` |
+| [memorize.md](prompts/memorize.md) | Memorize | Manages `.spec-lite/memory.md` — standing instructions for all agents. Use `/memorize bootstrap` to auto-generate. | `.spec-lite/memory.md` |
 | [orchestrator.md](prompts/orchestrator.md) | — | Meta-document: pipeline, memory protocol, conflict resolution | Reference only |
 
 ## Output Directory Structure
 
-spec-lite sub-agents produce artifacts in the `.spec/` directory (version-controlled project metadata):
+spec-lite sub-agents produce artifacts in the `.spec-lite/` directory (version-controlled project metadata):
 
 ```
-.spec/
+.spec-lite/
 ├── memory.md                  # Cross-cutting standards — authoritative source
 ├── brainstorm.md
 ├── plan.md                    # Default plan (simple projects) — user-modifiable
 ├── plan_<name>.md             # Named plans (complex projects, e.g., plan_order_management.md)
+├── architect_<name>.md        # Cloud & infrastructure architecture (e.g., architect_fintech_platform.md)
 ├── TODO.md                    # Enhancement backlog — maintained by planner + feature
 ├── features/
 │   ├── feature_user_management.md
@@ -184,7 +186,7 @@ See [orchestrator.md](prompts/orchestrator.md) for the complete workflow documen
 - The full sub-agent pipeline DAG
 - Memory protocol — which artifacts each sub-agent reads
 - Conflict resolution rules (user instruction > plan > sub-agent expertise)
-- Enhancement tracking via `.spec/TODO.md`
+- Enhancement tracking via `.spec-lite/TODO.md`
 - Invocation patterns for different scenarios (new project, feature addition, bug fix)
 
 ## CLI Commands
@@ -228,13 +230,13 @@ spec-lite relies on **git** for artifact versioning. When a plan or review is up
 
 spec-lite is designed to be forked and adapted:
 
-- **Bootstrap memory first** — run `/memorize bootstrap` after init to populate `.spec/memory.md` with your project's standards.
-- **Edit memory directly** — `.spec/memory.md` is the standing-instruction file. Your edits persist across all sub-agent invocations.
+- **Bootstrap memory first** — run `/memorize bootstrap` after init to populate `.spec-lite/memory.md` with your project's standards.
+- **Edit memory directly** — `.spec-lite/memory.md` is the standing-instruction file. Your edits persist across all sub-agent invocations.
 - **Add project-specific conventions** to the Project Context blocks or directly to memory.
 - **Remove sub-agents** you don't need.
 - **Add new sub-agents** following the same pattern (Persona → Required Context → Process → Output Template → Constraints).
 - **Modify output paths** to match your project's directory structure.
-- **Edit the plan** — `.spec/plan.md` (or `.spec/plan_<name>.md` for named plans) is a living document. Your edits take priority over sub-agent defaults.
+- **Edit the plan** — `.spec-lite/plan.md` (or `.spec-lite/plan_<name>.md` for named plans) is a living document. Your edits take priority over sub-agent defaults.
 - **Add stack snippets** — drop a `<language>.md` file into `src/stacks/` to add best-practice snippets for additional languages.
 
 Contributions welcome — especially for new sub-agent types, improvements to existing prompts, and real-world usage feedback.

package/dist/index.js

@@ -48,6 +48,13 @@ var CopilotProvider = class {
     }
     return existing;
   }
+  async getMemorySeedSource(workspaceRoot) {
+    const p = path.join(workspaceRoot, ".github", "copilot-instructions.md");
+    if (await fs.pathExists(p)) {
+      return { path: ".github/copilot-instructions.md", label: "GitHub Copilot global instructions" };
+    }
+    return null;
+  }
   getPostInitMessage() {
     return [
       "",
@@ -65,6 +72,42 @@ var CopilotProvider = class {
     ].join("\n");
   }
 };
+var SPEC_LITE_MARKER_START = "<!-- spec-lite:start -->";
+var SPEC_LITE_MARKER_END = "<!-- spec-lite:end -->";
+function generateSpecLiteBlock(installedPrompts) {
+  const lines = [
+    SPEC_LITE_MARKER_START,
+    "## spec-lite Sub-Agents",
+    "",
+    "This project uses [spec-lite](https://github.com/ranjithab/spec-lite) sub-agent prompts",
+    "for structured software engineering workflows.",
+    "",
+    "The following specialist sub-agents are available in `.github/copilot/`:",
+    ""
+  ];
+  for (const name of installedPrompts) {
+    lines.push(`- [${name}](.github/copilot/${name}.prompt.md)`);
+  }
+  lines.push(
+    "",
+    "To invoke a sub-agent in Copilot Chat, use the `#` file reference or type `/` to browse prompt files.",
+    SPEC_LITE_MARKER_END
+  );
+  return lines.join("\n");
+}
+function mergeCopilotInstructions(existingContent, installedPrompts) {
+  const block = generateSpecLiteBlock(installedPrompts);
+  if (!existingContent) {
+    return block + "\n";
+  }
+  const startIdx = existingContent.indexOf(SPEC_LITE_MARKER_START);
+  const endIdx = existingContent.indexOf(SPEC_LITE_MARKER_END);
+  if (startIdx !== -1 && endIdx !== -1 && endIdx > startIdx) {
+    return existingContent.slice(0, startIdx) + block + existingContent.slice(endIdx + SPEC_LITE_MARKER_END.length);
+  }
+  const separator = existingContent.endsWith("\n") ? "\n" : "\n\n";
+  return existingContent + separator + block + "\n";
+}
 
 // src/providers/claude-code.ts
 import path2 from "path";
@@ -105,6 +148,13 @@ var ClaudeCodeProvider = class {
     }
     return existing;
   }
+  async getMemorySeedSource(workspaceRoot) {
+    const p = path2.join(workspaceRoot, "CLAUDE.md");
+    if (await fs2.pathExists(p)) {
+      return { path: "CLAUDE.md", label: "Claude root instructions (CLAUDE.md)" };
+    }
+    return null;
+  }
   getPostInitMessage() {
     return [
       "",
@@ -150,10 +200,10 @@ function generateClaudeRootMd(installedPrompts) {
     "",
     "## Output Directory",
     "",
-    "Sub-agent outputs are written to the `.spec/` directory:",
+    "Sub-agent outputs are written to the `.spec-lite/` directory:",
     "",
     "```text",
-    ".spec/",
+    ".spec-lite/",
     "\u251C\u2500\u2500 brainstorm.md",
     "\u251C\u2500\u2500 plan.md                    # Default plan (simple projects)",
     "\u251C\u2500\u2500 plan_<name>.md              # Named plans (complex projects)",
@@ -197,6 +247,9 @@ var GenericProvider = class {
     }
     return existing;
   }
+  async getMemorySeedSource(_workspaceRoot) {
+    return null;
+  }
   getPostInitMessage() {
     return [
       "",
@@ -245,17 +298,17 @@ var PROMPT_CATALOG = {
   brainstorm: {
     title: "Brainstorm",
     description: "Refines a vague idea into a clear, actionable vision",
-    output: ".spec/brainstorm.md"
+    output: ".spec-lite/brainstorm.md"
   },
   planner: {
     title: "Planner",
     description: "Creates a detailed technical blueprint from requirements",
-    output: ".spec/plan.md or .spec/plan_<name>.md"
+    output: ".spec-lite/plan.md or .spec-lite/plan_<name>.md"
   },
   feature: {
     title: "Feature",
     description: "Breaks one feature into granular, verifiable vertical slices",
-    output: ".spec/features/feature_<name>.md"
+    output: ".spec-lite/features/feature_<name>.md"
   },
   implement: {
     title: "Implement",
@@ -265,17 +318,17 @@ var PROMPT_CATALOG = {
   code_review: {
     title: "Code Review",
     description: "Reviews code for correctness, architecture, and readability",
-    output: ".spec/reviews/code_review_<name>.md"
+    output: ".spec-lite/reviews/code_review_<name>.md"
   },
   security_audit: {
     title: "Security Audit",
     description: "Scans for vulnerabilities, misconfigurations, and security risks",
-    output: ".spec/reviews/security_audit_<scope>.md"
+    output: ".spec-lite/reviews/security_audit_<scope>.md"
   },
   performance_review: {
     title: "Performance Review",
     description: "Identifies bottlenecks and optimization opportunities",
-    output: ".spec/reviews/performance_review_<scope>.md"
+    output: ".spec-lite/reviews/performance_review_<scope>.md"
   },
   integration_tests: {
     title: "Integration Tests",
@@ -285,7 +338,7 @@ var PROMPT_CATALOG = {
   unit_tests: {
     title: "Unit Tests",
     description: "Generates comprehensive unit tests with edge-case coverage and smart coverage exclusions",
-    output: ".spec/features/unit_tests_<name>.md"
+    output: ".spec-lite/features/unit_tests_<name>.md"
   },
   devops: {
     title: "DevOps",
@@ -300,7 +353,7 @@ var PROMPT_CATALOG = {
   memorize: {
     title: "Memorize",
     description: "Stores standing instructions that all sub-agents enforce. Use `/memorize bootstrap` to auto-generate from project analysis.",
-    output: ".spec/memory.md"
+    output: ".spec-lite/memory.md"
   },
   technical_docs: {
     title: "Technical Docs",
@@ -311,6 +364,11 @@ var PROMPT_CATALOG = {
     title: "README",
     description: "Writes the project README and optional user guide",
     output: "README.md + docs/user_guide.md"
+  },
+  architect: {
+    title: "Architect",
+    description: "Designs cloud infrastructure, database strategy, and scaling architecture with Mermaid diagrams",
+    output: ".spec-lite/architect_<name>.md"
   }
 };
 var SKIP_FILES = /* @__PURE__ */ new Set(["orchestrator"]);
@@ -492,6 +550,34 @@ function buildProjectContextBlock(profile) {
   lines.push("");
   return lines.join("\n");
 }
+function buildSeededMemory(sourceContent, sourcePath, version) {
+  const date = (/* @__PURE__ */ new Date()).toISOString().split("T")[0];
+  return [
+    `<!-- Generated by spec-lite v${version} | sub-agent: memorize | seeded-from: ${sourcePath} | updated: ${date} -->`,
+    "",
+    "# Memory \u2014 Standing Instructions",
+    "",
+    `> These instructions were **auto-seeded** from \`${sourcePath}\` during \`spec-lite init\`.`,
+    "> They have not yet been organized into sections.",
+    "> Run `/memorize bootstrap` in your AI assistant to review, reorganize, and refine them.",
+    ">",
+    "> Memory is the **authoritative source** for coding standards, architecture, testing, logging, and security.",
+    "> Plans may contain plan-specific overrides but should not duplicate these rules.",
+    "> Managed by the Memorize sub-agent. Do not edit section headers manually.",
+    "> To add or change instructions, invoke: `/memorize <your instructions>`",
+    "> To override: `/memorize override <your instructions>`",
+    "> To generate from project analysis: `/memorize bootstrap`",
+    "",
+    `<!-- seed-start: raw content imported from ${sourcePath} -->`,
+    "",
+    `## Imported from \`${sourcePath}\``,
+    "",
+    sourceContent.trim(),
+    "",
+    "<!-- seed-end -->",
+    ""
+  ].join("\n");
+}
 async function initCommand(options) {
   const cwd = process.cwd();
   console.log(chalk.bold("\n\u26A1 spec-lite init\n"));
@@ -532,7 +618,16 @@ async function initCommand(options) {
   if (exclude.length > 0) {
     console.log(chalk.dim(`  Excluding: ${exclude.join(", ")}`));
   }
+  const memorySeedSource = await provider.getMemorySeedSource(cwd);
+  let preSeedContent = null;
+  if (memorySeedSource) {
+    const seedAbsPath = path6.join(cwd, memorySeedSource.path);
+    if (await fs6.pathExists(seedAbsPath)) {
+      preSeedContent = await fs6.readFile(seedAbsPath, "utf-8");
+    }
+  }
   const existingFiles = await provider.detectExisting(cwd);
+  let globalAction = null;
   if (existingFiles.length > 0 && !options.force) {
     console.log(
       chalk.yellow(
@@ -563,7 +658,8 @@ ${existingFiles.map((f) => `    - ${f}`).join("\n")}`
       console.log(chalk.dim("  Aborted."));
       return;
     }
-    if (answer.action === "skip") {
+    globalAction = answer.action;
+    if (globalAction === "skip") {
       console.log(chalk.dim("  Skipping existing files."));
     }
   }
@@ -575,23 +671,10 @@ ${existingFiles.map((f) => `    - ${f}`).join("\n")}`
   for (const prompt of prompts) {
     const targetRelPath = provider.getTargetPath(prompt.name);
     const targetAbsPath = path6.join(cwd, targetRelPath);
-    if (!options.force && existingFiles.includes(targetRelPath) && existingFiles.length > 0) {
-      const answer = await inquirer.prompt([
-        {
-          type: "list",
-          name: "action",
-          message: `How should we handle existing files?`,
-          choices: [
-            { name: "Overwrite", value: "overwrite" },
-            { name: "Skip", value: "skip" }
-          ]
-        }
-      ]);
-      if (answer.action === "skip") {
-        skipped++;
-        installedPrompts.push(prompt.name);
-        continue;
-      }
+    if (!options.force && existingFiles.includes(targetRelPath) && globalAction === "skip") {
+      skipped++;
+      installedPrompts.push(prompt.name);
+      continue;
     }
     let content = prompt.content;
     if (contextBlock) {
@@ -615,10 +698,22 @@ ${existingFiles.map((f) => `    - ${f}`).join("\n")}`
     console.log(chalk.green(`  \u2713 CLAUDE.md`));
     written++;
   }
+  if (provider.alias === "copilot") {
+    const copilotInstructionsPath = path6.join(cwd, ".github", "copilot-instructions.md");
+    await fs6.ensureDir(path6.join(cwd, ".github"));
+    const existingContent = await fs6.pathExists(copilotInstructionsPath) ? await fs6.readFile(copilotInstructionsPath, "utf-8") : null;
+    const merged = mergeCopilotInstructions(existingContent, installedPrompts);
+    await fs6.writeFile(copilotInstructionsPath, merged, "utf-8");
+    if (existingContent) {
+      console.log(chalk.green(`  \u2713 .github/copilot-instructions.md (updated with spec-lite block)`));
+    } else {
+      console.log(chalk.green(`  \u2713 .github/copilot-instructions.md`));
+    }
+    written++;
+  }
   const specDirs = [
-    ".spec",
-    path6.join(".spec", "features"),
-    path6.join(".spec", "reviews")
+    path6.join(".spec-lite", "features"),
+    path6.join(".spec-lite", "reviews")
   ];
   for (const dir of specDirs) {
     const absDir = path6.join(cwd, dir);
@@ -627,7 +722,7 @@ ${existingFiles.map((f) => `    - ${f}`).join("\n")}`
       console.log(chalk.green(`  \u2713 ${dir}/`));
     }
   }
-  const todoPath = path6.join(cwd, ".spec", "TODO.md");
+  const todoPath = path6.join(cwd, ".spec-lite", "TODO.md");
   if (!await fs6.pathExists(todoPath)) {
     const todoContent = [
       "# TODO \u2014 Enhancements & Ideas",
@@ -649,7 +744,7 @@ ${existingFiles.map((f) => `    - ${f}`).join("\n")}`
       ""
     ].join("\n");
     await fs6.writeFile(todoPath, todoContent, "utf-8");
-    console.log(chalk.green(`  \u2713 .spec/TODO.md`));
+    console.log(chalk.green(`  \u2713 .spec-lite/TODO.md`));
   }
   if (projectProfile) {
     const snippet = getStackSnippet(projectProfile.language);
@@ -674,6 +769,32 @@ ${existingFiles.map((f) => `    - ${f}`).join("\n")}`
       }
     }
   }
+  let memorySeedWritten = false;
+  const memoryPath = path6.join(cwd, ".spec-lite", "memory.md");
+  if (preSeedContent && memorySeedSource && !await fs6.pathExists(memoryPath)) {
+    console.log(
+      chalk.cyan(`
+  \u{1F4A1} Found existing ${memorySeedSource.label} (${memorySeedSource.path}).`)
+    );
+    const seedAnswer = await inquirer.prompt([
+      {
+        type: "confirm",
+        name: "seedMemory",
+        message: `Seed .spec-lite/memory.md from it so /memorize bootstrap can refine your existing conventions?`,
+        default: true
+      }
+    ]);
+    if (seedAnswer.seedMemory) {
+      const seedPkg = await loadPackageVersion();
+      const seededContent = buildSeededMemory(preSeedContent, memorySeedSource.path, seedPkg);
+      await fs6.ensureDir(path6.join(cwd, ".spec-lite"));
+      await fs6.writeFile(memoryPath, seededContent, "utf-8");
+      memorySeedWritten = true;
+      written++;
+      console.log(chalk.green(`  \u2713 .spec-lite/memory.md (seeded from ${memorySeedSource.path})`));
+      console.log(chalk.dim("     \u21B3 Run /memorize bootstrap to organize and refine into standing instructions"));
+    }
+  }
   const pkg2 = await loadPackageVersion();
   const config = {
     version: pkg2,
@@ -693,14 +814,24 @@ ${existingFiles.map((f) => `    - ${f}`).join("\n")}`
     )
   );
   console.log(provider.getPostInitMessage());
-  if (projectProfile) {
-    console.log(
-      chalk.cyan(
-        "\n  \u{1F4CC} Next step: Run "
-      ) + chalk.bold("/memorize bootstrap") + chalk.cyan(
-        " in your AI assistant to auto-generate\n     coding standards, architecture guidelines, and best practices\n     for your project based on the profile you just provided."
-      )
-    );
+  if (projectProfile || memorySeedWritten) {
+    if (memorySeedWritten) {
+      console.log(
+        chalk.cyan(
+          "\n  \u{1F4CC} Next step: Run "
+        ) + chalk.bold("/memorize bootstrap") + chalk.cyan(
+          " in your AI assistant.\n     It will see your seeded memory and offer to merge or refine it\n     into properly organized standing instructions."
+        )
+      );
+    } else {
+      console.log(
+        chalk.cyan(
+          "\n  \u{1F4CC} Next step: Run "
+        ) + chalk.bold("/memorize bootstrap") + chalk.cyan(
+          " in your AI assistant to auto-generate\n     coding standards, architecture guidelines, and best practices\n     for your project based on the profile you just provided."
+        )
+      );
+    }
   }
 }
 async function loadPackageVersion() {
@@ -710,7 +841,7 @@ async function loadPackageVersion() {
     const pkg2 = require3("../../package.json");
     return pkg2.version;
   } catch {
-    return "1.0.0";
+    return "0.0.3";
   }
 }
 
@@ -810,6 +941,14 @@ async function updateCommand(options) {
     await fs7.writeFile(claudeMdPath, claudeMdContent, "utf-8");
     console.log(chalk2.green(`  \u2713 CLAUDE.md (regenerated)`));
   }
+  if (provider.alias === "copilot") {
+    const copilotInstructionsPath = path7.join(cwd, ".github", "copilot-instructions.md");
+    await fs7.ensureDir(path7.join(cwd, ".github"));
+    const existingContent = await fs7.pathExists(copilotInstructionsPath) ? await fs7.readFile(copilotInstructionsPath, "utf-8") : null;
+    const merged = mergeCopilotInstructions(existingContent, config.installedPrompts);
+    await fs7.writeFile(copilotInstructionsPath, merged, "utf-8");
+    console.log(chalk2.green(`  \u2713 .github/copilot-instructions.md (updated)`));
+  }
   config.updatedAt = (/* @__PURE__ */ new Date()).toISOString();
   try {
     const { createRequire: createRequire2 } = await import("module");