paladini · paladini · Jul 15, 2026 · Jul 20, 2026 · Jul 20, 2026
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -5,27 +5,41 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [Unreleased]
 
-## [0.3.0] - 2026-03-23
+## [1.1.0] - 2026-07-15
 
 ### Added
 
-- **README banner** — `docs/assets/og-image.svg` now displayed as a full-width banner at the top of the README
-- **Agent instruction file templates** — `templates/AGENTS.md` (for Windsurf/Cascade, Cline, Codex, and other agents) and `templates/.github/copilot-instructions.md` (for GitHub Copilot) ship with the package so users can drop them into their projects with one `cp` command
-- **Agent Instructions Guide** — new `docs/ai-instructions.md` documents how to use `AGENTS.md` and `copilot-instructions.md` with mcp-me, including global VS Code setup, a before/after comparison table, and tips on combining profile context with project-specific rules
-- **README "Agent Instruction Files" section** — quick-start snippet and link to the guide added between "Configure Your AI Assistant" and "Profile Schema"
+- **Writing dimension** — `writing/style.yaml` format profiles and `writing/corpus/` for published texts
+- **Corpus sync** — Medium, Substack, Blogger, and DEV.to generators write full articles to `writing/corpus/`
+- **Writing MCP surface** — Resources `me://writing/style`, `me://writing/corpus`, `me://writing/samples`
+- **Writing tools** — `analyze_writing_style`, `search_writing_corpus`, `get_writing_references`
+- **Writing prompts** — `describe_my_writing`, `emulate_my_voice`, `rewrite_in_my_voice`
+- **CLI commands** — `mcp-me sync-corpus`, `mcp-me analyze-writing`, `--no-corpus` flag on generate
+- **Profile completeness** — `get_profile_completeness` tool with domain scores and generator suggestions
+- **Open Plugins** — `voice-writer` and `style-analyst` agents, `writing-voice` skill, `match-voice` rule, `/emulate-voice` command
 
-## [Unreleased]
+### Changed
+
+- **`ask_about_me`** — Description clarifies it returns context for the host LLM (not server-side Q&A)
+- **`docs/ai-instructions.md`** — Documents writing MCP tools, prompts, and `get_profile_completeness`
+
+## [1.0.0] - 2026-07-15
 
 ### Added
 
-- **Open Plugins structure** — Full Cursor/Open Plugins layout at repo root for auto-detect on cursor.directory: `.cursor-plugin/plugin.json`, `mcp.json`, `rules/`, `skills/`, `agents/`, `commands/`, `hooks/hooks.json`, `.lsp.json`
-- **Plugin validation** — `npm run validate:open-plugin` script and `tests/open-plugin-structure.test.ts`
+- **Stable API declaration** — 8 core YAML schemas, MCP core surface, and `McpMePlugin` interface declared stable
+- **Identity fields** — Optional `headline` and `physical` (opt-in) in `identity.yaml`
+- **Schema fix** — `interests.topics` officially supported (generators already wrote this field)
+- **Documentation** — `docs/identity-model.md`, `docs/generator-tiers.md`, `docs/migration-1.0.md`
+- **Integration tests** — CLI, MCP server, and config-generate test suites
+- **Plugin tests** — Behavioral tests for devto and wakatime plugins
 
 ### Changed
 
-- **Directory submissions** — Opened Chrome handoff for mcp.so, Smithery, and Cursor Directory; built `mcp-me.mcpb` for Smithery; pushed wong2/awesome-mcp-servers PR branch; verified PulseMCP not yet indexed (registry active since 2026-07-10)
-- **Cursor Directory** — Submission path updated to `cursor.directory/plugins/new` Auto (GitHub) now that Open Plugins components are in the repo
+- **Plugin config docs** — Clarified `plugins.yaml` is legacy; use `.mcp-me.yaml`
+- **Spotify plugin docs** — Aligned tool/resource names with implementation
 
 ## [0.6.0] - 2026-07-09
 

diff --git a/README.md b/README.md
@@ -6,7 +6,9 @@
 
 **Your AI assistants don't know who you are.** Every time you start a conversation with Claude, Copilot, Cursor, or Windsurf, it's a blank slate — no context about your skills, your projects, your career, or what you care about.
 
-**mcp-me fixes that.** It creates a structured personal profile that any AI assistant can read via the [Model Context Protocol (MCP)](https://modelcontextprotocol.io/). Think of it as a **digital identity layer for AI** — your bio, career, skills, interests, projects, and more, always available to every AI tool you use.
+**mcp-me fixes that.** It creates a structured personal profile that any AI assistant can read via the [Model Context Protocol (MCP)](https://modelcontextprotocol.io/). Think of it as a **complete digital identity layer for AI** — your bio, career, skills, interests, projects, writing voice, and more, always available to every AI tool you use.
+
+See [Identity Model](docs/identity-model.md) for the full architecture vision.
 
 ```
 You: "Write me a cover letter for this job"
@@ -17,7 +19,7 @@ AI:  (reads your me://career, me://skills, me://projects)
 
 ## Why mcp-me?
 
-- **AI that knows you** — Your assistants remember your skills, career, projects, and personality across every conversation
+- **AI that knows you** — Your assistants remember your skills, career, projects, personality, and writing style across every conversation
 - **Auto-generated** — Pull data from 329 registered generators (implemented across 44 generator source files) with one command
 - **Privacy-first** — All data stays local in YAML files on your machine. Nothing is sent to any cloud.
 - **Real-time plugins** — 13 live integrations (Spotify now playing, GitHub repos, Last.fm scrobbles) that AI queries on demand
@@ -126,6 +128,10 @@ All commands work with `npx` (zero install) or with `mcp-me` directly if install
   personality.yaml     ← Your data (traits, values)
   goals.yaml           ← Your data (short/long-term)
   faq.yaml             ← Your data (Q&A pairs)
+  writing/             ← Writing voice (1.1+)
+    style.yaml         ← Format profiles and tone
+    corpus/            ← Your published texts (.md)
+      _manifest.yaml
 ```
 
 ## Configure Your AI Assistant

diff --git a/agents/style-analyst.md b/agents/style-analyst.md
@@ -0,0 +1,22 @@
+---
+name: style-analyst
+description: Analyze and compare the user's writing style across format profiles.
+---
+
+# Style Analyst Agent
+
+Use when the user wants to understand how they write or compare different writing contexts.
+
+## Workflow
+
+1. Read `me://writing/style` for defined format profiles
+2. Call `analyze_writing_style` with optional profile filter
+3. Compare profiles if requested (personal_blog vs tech_news)
+4. Summarize: tone, sentence length, vocabulary, first-person usage, punctuation patterns
+5. Suggest which profile fits a given writing task
+
+## Output format
+
+- **Overall voice** — shared traits across profiles
+- **Per-profile** — how blog differs from news differs from social
+- **Recommendations** — which profile to use for a given task
diff --git a/agents/voice-writer.md b/agents/voice-writer.md
@@ -0,0 +1,24 @@
+---
+name: voice-writer
+description: Write text in the user's authentic voice using mcp-me writing corpus and format profiles.
+---
+
+# Voice Writer Agent
+
+Use this agent when the user wants content written in their own voice.
+
+## Workflow
+
+1. Read `me://writing/style` to see format profiles (personal_blog, tech_news, ironic_thread)
+2. Ask or infer which format profile fits the request
+3. Call `get_writing_references` with the topic and profile
+4. Read `me://writing/samples?profile=<name>` for tone examples
+5. Load factual context from `me://career`, `me://skills`, `me://projects` as needed
+6. Use the `emulate_my_voice` prompt with topic, profile, and references
+7. Produce authentic text — never generic filler
+
+## Rules
+
+- Match the selected format profile's tone and length
+- Cite real experiences from the profile when relevant
+- Do not mention that you are emulating — write naturally
diff --git a/commands/emulate-voice.md b/commands/emulate-voice.md
@@ -0,0 +1,23 @@
+# Emulate Voice
+
+Write content in the user's authentic voice using mcp-me writing corpus.
+
+## Usage
+
+```
+/emulate-voice <topic> [--profile personal_blog|tech_news|ironic_thread]
+```
+
+## Steps
+
+1. Load writing style from `me://writing/style`
+2. Get references via `get_writing_references(topic, profile)`
+3. Read samples from `me://writing/samples?profile=<profile>`
+4. Invoke `emulate_my_voice` prompt with topic and profile
+5. Output the generated text
+
+## Examples
+
+- `/emulate-voice Rust async patterns --profile personal_blog`
+- `/emulate-voice AI regulation --profile tech_news`
+- `/emulate-voice hot take on tabs vs spaces --profile ironic_thread`
diff --git a/docs/ai-instructions.md b/docs/ai-instructions.md
@@ -100,6 +100,24 @@ documentation, career questions, open-source contributions, etc.
 | `me://personality` | Traits, values, work style |
 | `me://goals` | Short and long-term goals |
 | `me://faq` | Pre-answered facts about me |
+| `me://writing/style` | Writing format profiles and voice guide (1.1+) |
+| `me://writing/corpus` | Published texts manifest (1.1+) |
+| `me://writing/samples` | Writing excerpts by format profile (1.1+) |
+
+## Writing and voice (1.1+)
+
+For writing tasks, also use these MCP tools and prompts:
+
+| Tool / Prompt | When to use |
+|---|---|
+| `get_writing_references(topic, profile)` | Find corpus excerpts before drafting |
+| `analyze_writing_style(profile?)` | Stats on tone, sentence length, vocabulary |
+| `search_writing_corpus(query)` | Keyword search across local `.md` files |
+| `emulate_my_voice` prompt | Generate text in my voice for a topic |
+| `rewrite_in_my_voice` prompt | Rewrite text to match my style |
+| `get_profile_completeness` | See what's missing from my profile |
+
+**Note:** `ask_about_me` returns structured profile context for the host LLM to interpret — mcp-me does not run an LLM server-side.
 
 ## When to proactively load my profile
 

diff --git a/docs/creating-plugins.md b/docs/creating-plugins.md
@@ -293,7 +293,7 @@ Every registered built-in plugin is automatically tested by `tests/plugins/plugi
 | Plugin | Resources | Tools | Auth |
 |---|---|---|---|
 | **github** | profile, repos, activity, languages | `get_github_repos` | Optional token |
-| **spotify** | now playing, top artists, top tracks, recently played | `get_spotify_top`, `get_spotify_now_playing` | OAuth (refresh token) |
+| **spotify** | top artists, top tracks, recently played | `get_spotify_now_playing`, `get_spotify_music_taste` | OAuth (refresh token) |
 | **linkedin** | profile, experience, education, skills | `search_linkedin_data` | Data export JSON |
 | **wakatime** | stats, languages, activity | `get_wakatime_stats` | Optional API key |
 | **devto** | profile, articles | `get_devto_articles` | Optional API key |

diff --git a/docs/generator-tiers.md b/docs/generator-tiers.md
@@ -0,0 +1,73 @@
+# Generator Tiers
+
+mcp-me registers **329 generators**. Not all fetch live data — this document explains what each tier means so you can set honest expectations.
+
+## Tiers
+
+| Tier | Definition | What you get |
+|------|------------|--------------|
+| **Verified** | Real public API + behavioral tests | Rich, accurate profile data |
+| **Partial** | Real API, structural tests only | Working data, less test coverage |
+| **Static** | User input → placeholder link + FAQ | Social link and a FAQ entry, no fetch |
+
+## Verified generators
+
+These have real API integration and behavioral tests:
+
+| Generator | Category | Data populated |
+|-----------|----------|----------------|
+| `github` | code | identity, skills, projects, career, interests, faq |
+| `gitlab` | code | identity, skills, projects, faq |
+| `npm` | packages | projects, faq |
+| `devto` | writing | identity, projects, skills, faq |
+| `medium` | writing | identity, projects, skills, interests, faq |
+| `substack` | writing | identity, projects, faq |
+| `blogger-backup` | writing | projects, faq (from XML export) |
+| `hashnode` | writing | identity, projects, faq |
+| `wordpress` | writing | projects, faq |
+| `linkedin` | identity | identity, career, skills, faq (JSON export) |
+| `goodreads` | interests | interests, faq |
+| `stackoverflow` | code | skills, faq |
+| `orcid` | learning | career, projects, faq |
+| `semantic-scholar` | learning | projects, faq |
+| `openlibrary` | writing | projects, interests, faq |
+| `youtube` | writing | projects, faq |
+
+## Partial generators
+
+Real APIs with structural/harness tests:
+
+| Generator | Notes |
+|-----------|-------|
+| `bluesky` | Public AT Protocol API |
+| `reddit` | Public JSON endpoints |
+| `mastodon` | Public API |
+| `hackernews` | Firebase API |
+| `npm` | Registry API |
+| `wakatime` | Requires API key |
+| `gravatar` | Email hash lookup |
+| `keybase` | Public API |
+
+## Static generators (~284)
+
+Batch generators in `batch-*.ts` files accept a username or ID and produce:
+
+- A social link in `identity.contact.social`
+- A FAQ entry describing presence on that platform
+- Sometimes `interests.topics` tags
+
+They do **not** call external APIs. They are useful for documenting your presence across many platforms when you don't need live data.
+
+Examples: `--notion`, `--obsidian`, `--duolingo`, `--peloton`, `--zodiac`, `--mbti`
+
+## Writing generators and corpus (1.1+)
+
+Verified writing generators (Medium, Substack, Blogger, DEV.to) also sync full article text to `writing/corpus/` when corpus sync is enabled (default). Use `--no-corpus` to opt out.
+
+## Choosing generators
+
+1. Start with **Verified** sources you actually use: `--github`, `--devto`, `--medium`
+2. Add **Partial** sources for live social presence
+3. Use **Static** generators only to document platform links — not for rich data
+
+Run `mcp-me generate --help` to see all available flags grouped by category.
diff --git a/docs/identity-model.md b/docs/identity-model.md
@@ -0,0 +1,80 @@
+# Identity Model
+
+mcp-me is a **complete digital identity layer** for any AI assistant connected via MCP. It aggregates who you are, what you've done, what you know, what you care about, and how you write — all from local YAML files and optional live plugins.
+
+## North star
+
+> **mcp-me transforms any AI from a generic assistant into one that knows you.**
+
+When connected to Claude, Copilot, Cursor, Windsurf, or any MCP client, the agent can answer questions and produce content grounded in your real profile — not generic placeholders.
+
+## Identity domains
+
+| Domain | YAML / Resource | What it covers |
+|--------|-----------------|----------------|
+| **Who I am** | `identity.yaml` → `me://identity` | Name, bio, headline, location, contact, social links, optional physical attributes |
+| **Career** | `career.yaml` → `me://career` | Work experience, education, certifications |
+| **Skills** | `skills.yaml` → `me://skills` | Technical/soft skills, tools, programming languages |
+| **Projects** | `projects.yaml` → `me://projects` | Open-source, articles, packages, portfolio |
+| **Interests** | `interests.yaml` → `me://interests` | Hobbies, topics, music, books, travel |
+| **Personality** | `personality.yaml` → `me://personality` | Traits, values, work style, motivations |
+| **Goals** | `goals.yaml` → `me://goals` | Short/medium/long-term aspirations |
+| **FAQ** | `faq.yaml` → `me://faq` | Pre-answered facts — persistent memory |
+| **Writing** | `writing/` → `me://writing/*` | Style profiles, corpus of real texts (1.1+) |
+| **Live presence** | Plugins → `me://github/*`, etc. | Real-time data (Spotify, GitHub, Last.fm…) |
+
+Writing is a **first-class dimension** but never isolated — emulated text should cite real experiences, projects, and opinions from the rest of the profile.
+
+## Data flow
+
+```
+Sources (APIs, exports, manual YAML, local corpus)
+        ↓
+  Generators (batch)  +  Manual edits
+        ↓
+  Profile directory (~/.mcp-me)
+        ↓
+  mcp-me serve  →  MCP resources, tools, prompts
+        ↓
+  Any MCP client (agent-agnostic)
+```
+
+### Generators vs plugins
+
+| | Generators | Plugins |
+|---|-----------|---------|
+| **When** | One-time (`mcp-me generate`) | Live during `mcp-me serve` |
+| **Output** | Static YAML files | MCP resources/tools on demand |
+| **Auth** | Usually none (public APIs) | Sometimes OAuth (Spotify) |
+
+## Agent-agnostic design
+
+Core value lives in the **MCP server** — not in Cursor-specific rules or skills. Open Plugins (rules, skills, agents, commands) are **accelerators** for Cursor users, not requirements.
+
+### Core MCP surface
+
+**Resources:** `me://identity`, `me://career`, `me://skills`, `me://interests`, `me://personality`, `me://goals`, `me://projects`, `me://faq`, plus `me://writing/*` (1.1+)
+
+**Tools:**
+- `ask_about_me` — Returns structured profile context for the host LLM to interpret (not server-side Q&A)
+- `search_profile` — Keyword search across all profile data
+- Writing tools (1.1+): `analyze_writing_style`, `search_writing_corpus`, `get_writing_references`
+- `get_profile_completeness` (1.2+) — Profile fill percentage and generator suggestions
+
+**Prompts:** `introduce_me`, `summarize_career`, `technical_profile`, `collaboration_fit`, plus writing prompts (1.1+)
+
+## Privacy
+
+- All profile data stays **local** on your machine
+- Sensitive fields (`identity.physical`, etc.) are **opt-in only** — never populated by generators
+- Secrets use `_env` suffix keys in `.mcp-me.yaml` — never stored in YAML profile files
+
+## Stable API (1.0+)
+
+The following are declared stable as of mcp-me 1.0.0:
+
+- 8 core YAML schemas (`identity`, `career`, `skills`, `interests`, `personality`, `goals`, `projects`, `faq`)
+- Core MCP resources, tools, and prompts listed above
+- `McpMePlugin` interface in `src/plugin-engine/types.ts`
+
+See [Migration Guide](migration-1.0.md) for upgrading from 0.x.
diff --git a/docs/index.html b/docs/index.html
@@ -64,7 +64,7 @@
     "license": "https://opensource.org/licenses/MIT",
     "codeRepository": "https://github.com/paladini/mcp-me",
     "programmingLanguage": "TypeScript",
-    "softwareVersion": "0.6.0"
+    "softwareVersion": "1.1.0"
   }
   </script>