@flydocs/cli 0.5.0-beta.1 → 0.5.0-beta.10

package/template/CHANGELOG.md

@@ -7,9 +7,83 @@ Versioning: [Semantic Versioning](https://semver.org/).
 
 ---
 
+## [0.5.0-beta.10] — 2026-02-22
+
+### Fixed
+
+- **`.env.example` not updating** — reclassified from `template_files` to `owned_files`
+  in manifest. Now overwritten on update so new config entries (like `CONTEXT7_API_KEY`)
+  and copy fixes propagate to existing installations.
+
+---
+
+## [0.5.0-beta.9] — 2026-02-22
+
+### Changed
+
+- **Banner copy** — title changed from "FlyDocs Core" to "FlyDocs (Beta)",
+  subtitle changed to "Spec-Driven Context That Helps Engineers Fly"
+- **Beta CTA** — "priority support" replaced with "web portal", URL now
+  renders as clickable `flydocs.ai` hyperlink via OSC 8 terminal escape
+
+---
+
+## [0.5.0-beta.8] — 2026-02-22
+
+### Added
+
+- **Context7 documentation skill** (`flydocs-context7`) — stdlib-only Python script
+  that calls the Context7 REST API to fetch up-to-date library documentation.
+  Replaces the removed Context7 MCP with a skill+script pattern.
+  - `context7.py search <library> [query]` — resolve library name to Context7 ID
+  - `context7.py docs <library_id> <query>` — fetch relevant documentation snippets
+  - Works without API key (anonymous, ~1,000 calls/month); optional
+    `CONTEXT7_API_KEY` in `.env` for higher rate limits
+  - Auto-approved by hook for seamless agent invocation
+  - Cursor rule included for cross-IDE support
+- **Context7 API key** in `.env.example` — optional configuration for higher rate limits
+
+### Fixed
+
+- `.env.example` — corrected "Figma MCP" reference to "flydocs-figma skill"
+
+## [0.5.0-beta.7] — 2026-02-22
+
+### Removed
+
+- **All MCP references** — removed `.cursor/mcp.json`, `prefer-scripts.py` hook,
+  MCP config preservation in update flow, MCP section in instructions template.
+  Skills fully replace MCPs as the FlyDocs architecture.
+- **Phantom skills** — removed references to `implementation-flow`, `review-workflow`,
+  `spec-templates` skills that were never shipped.
+
+### Fixed
+
+- **Command injection** (`connect.ts`) — replaced `execSync` subprocess with native
+  `fetch()` for Linear API key validation.
+- **Version comparison** — pre-release versions (e.g. `0.5.0-beta.3`) now compare
+  correctly instead of returning NaN.
+- **Auto-approve hook** — tightened regex to anchor script paths, preventing potential
+  pattern-matching bypasses.
+- **Post-edit hook** — added path traversal validation using `os.path.realpath()`.
+- **Post-install scripts** — switched from `execSync` to `execFileSync` for subprocess safety.
+
+### Changed
+
+- **Stack detection** — now reads `flydocs/context/project.md` in addition to
+  `package.json` for declared stack frameworks.
+- **Community skills on update** — filters already-installed skills before prompting.
+- **Cloud exposure** — `.env.example` marked as cloud tier only, `flydocs/README.md`
+  rewritten as tier-generic, config defaults changed to `tier: "local"`.
+- **Deprecated cleanup** — `.cursor/mcp.json` and `prefer-scripts.py` added to
+  deprecated files list for automatic cleanup on update.
+
+---
+
 ## [0.4.0] — 2026-02-18
 
 ### Added
+
 - **Context Graph skill** (`flydocs-context-graph`) — relationship-aware knowledge
   graph that connects skills, ADRs, issues, modules, and sessions. Implemented as
   ADR-006 across five phases:
@@ -32,12 +106,14 @@ Versioning: [Semantic Versioning](https://semver.org/).
   graph automatically.
 
 ### Changed
+
 - **Session wrap procedure** — `session.md` now includes a step to record session
   outcomes in the context graph via `graph_session.py`. Skips silently if not installed.
 - **Session staleness policy** — session nodes within 7 days get full weight, linear
   decay to 30 days, then excluded from context.
 
 ### Migration
+
 - **`.gitignore` update** — `flydocs/context/graph.json` is automatically added to
   `.gitignore` during update if missing. The graph is gitignored because it contains
   session-specific data and can be rebuilt from sources.
@@ -47,6 +123,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.10] — 2026-02-17
 
 ### Fixed
+
 - **API null data crash** (FLY-178) — `flydocs_api.py` now normalizes `null` data
   responses to empty dict before returning. Fixes `AttributeError` crashes in all
   20+ scripts when Linear returns API errors. `create_issue.py` also retries without
@@ -65,6 +142,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.9] — 2026-02-17
 
 ### Added
+
 - **Skill search and install CLI** (ENG-443) — `flydocs skills` subcommand with
   `search`, `add`, `remove`, and `list` operations. Search the curated community
   skills catalog, install from known names or GitHub repo references, and manage
@@ -81,6 +159,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.8] — 2026-02-14
 
 ### Added
+
 - **`create_issue.py --description-file` and stdin** — both cloud and local tiers now
   accept `--description-file PATH` or piped stdin for issue descriptions. Eliminates
   the two-step create + update_description workaround for long descriptions.
@@ -92,6 +171,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.7] — 2026-02-11
 
 ### Added
+
 - **Skill manifest generator** (ENG-442) — `generate_manifest.py` scans SKILL.md
   frontmatter and produces a compressed skills index table. Injected into CLAUDE.md
   and AGENTS.md between `<!-- flydocs:skills-manifest -->` markers. Bridges skill
@@ -104,12 +184,14 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.6] — 2026-02-10
 
 ### Added
+
 - **Milestone, project, and dueDate in issue output** — `list_issues.py` now returns
   `milestone`, `milestoneId`, `milestoneSortOrder`, `project`, `projectId`, and `dueDate`
   for every issue. `get_issue.py` returns the same fields (minus `milestoneSortOrder`).
   Enables milestone-grouped dashboards and priority-based session planning.
 
 ### Changed
+
 - **Session start workflow rewritten** — `session.md` now instructs agents to group issues
   by milestone (in milestone order), identify the current milestone, suggest a starting point
   based on priority cascade (blocked → in-progress → due soon → highest priority in current
@@ -117,6 +199,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
   the product scope.
 
 ### Fixed
+
 - **`create_milestone.py` missing projectId** — `ProjectMilestoneCreateInput` requires a
   `projectId`. Script now defaults to first `activeProject` from config, with `--project`
   flag to override.
@@ -126,10 +209,12 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.5] — 2026-02-10
 
 ### Added
+
 - **`list_projects.py` product scope** (ENG-475) — now respects `workspace.activeProjects`
   config. `--all` bypasses product scope to show all team projects.
 
 ### Fixed
+
 - **`list_projects.py` 400 error** (ENG-474) — state filter was using nested
   `{type: {nin: [...]}}` instead of direct `StringComparator` `{nin: [...]}`.
   Both `--active` (started only) and default (excludes completed/canceled) now work.
@@ -143,6 +228,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.4] — 2026-02-10
 
 ### Added
+
 - **`list_issues.py --milestone` filter** — filter issues by project milestone ID.
   Get milestone IDs from `list_milestones.py`, then pass to `list_issues.py --milestone <ID>`
   to see all issues within a specific milestone.
@@ -152,6 +238,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.3] — 2026-02-10
 
 ### Fixed
+
 - **Shell escaping in `update_description.py`** — descriptions with apostrophes,
   parentheses, quotes, and other shell metacharacters no longer break when passed
   via `--text`. Scripts now accept stdin as input, allowing safe heredoc piping:
@@ -164,6 +251,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.2] — 2026-02-09
 
 ### Fixed
+
 - **API connection reliability** — reduced socket timeout from 30s to 10s and increased
   max retries from 3 to 5. Resolves intermittent hangs caused by DNS returning unreachable
   Cloudflare edge IPs. Scripts now fail over to a working IP within seconds instead of
@@ -176,11 +264,13 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.1] — 2026-02-08
 
 ### Fixed
+
 - **`list_issues.py` product scope** — scoping via `activeProjects` and `product.labelIds`
   now applies regardless of `--active` or `--status` flags. Previously, these flags
   bypassed product scope entirely, returning all team issues.
 
 ### Added
+
 - **`build_product_scope()` API helper** — new method on `FlyDocsClient` implements the
   product scope cascade: `activeProjects` → `product.labelIds` → team-wide.
 - **Multiple `activeProjects` support** — `list_issues.py` now uses OR logic when
@@ -193,6 +283,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.3.0] — 2026-02-08
 
 ### Added
+
 - **`update_issue.py` bulk update script** — set multiple fields (priority, estimate,
   assignee, state, description, comment) in a single API call instead of separate scripts.
 - **`list_issues.py --active` flag** — returns all non-terminal issues (excludes Done,
@@ -204,12 +295,14 @@ Versioning: [Semantic Versioning](https://semver.org/).
   cached within each script invocation to reduce redundant API calls.
 
 ### Changed
+
 - **`transition.py` optimized** — combined state update and cycle assignment into a
   single `issueUpdate` mutation (was two separate calls for IMPLEMENTING transitions).
 - **Session start efficiency** — single `list_issues.py --active` call replaces
   per-status queries. Issues grouped by status in response.
 
 ### Removed
+
 - **`status_summary.py`** — retired from cloud mechanism. Use `list_issues.py --active`
   and group results by the status field instead.
 
@@ -218,6 +311,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.2.1] — 2026-02-07
 
 ### Fixed
+
 - **Commands not fully populating** — changed from additive `cp` to clear-then-copy
   for `.claude/commands/` and `.cursor/commands/` directories.
 - **`docflow-core.mdc` not cleaned up** — added to deprecated cursor rules list.
@@ -227,6 +321,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
   synced with version file.
 
 ### Added
+
 - **`--force` flag** — force update even when versions match.
 - **Old cursor rule directories and commands** added to deprecated cleanup.
 
@@ -235,6 +330,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.2.0] — 2026-02-06
 
 ### Added
+
 - **Scenario-aware setup command** — `/flydocs-setup` rewritten as a comprehensive
   agent prompt handling three scenarios: New Project, DocFlow Migration, FlyDocs Update.
   Includes proactive repo scanning for PRDs and existing documentation.
@@ -247,10 +343,12 @@ Versioning: [Semantic Versioning](https://semver.org/).
   pre-v1.0 artifacts: old skills, rules, hooks, and cursor rules.
 
 ### Changed
+
 - **Version file synced with manifest** — `.flydocs/version` now matches
   `manifest.json` version field (was out of sync in 0.1.0).
 
 ### Removed
+
 - Orphaned hook scripts: `session-end.py`, `linear-auto-approve.py`
 
 ---
@@ -258,6 +356,7 @@ Versioning: [Semantic Versioning](https://semver.org/).
 ## [0.1.0] — 2026-01-15
 
 ### Added
+
 - Initial release of FlyDocs Core
 - Workflow skill (`flydocs-workflow`) with full lifecycle stages
 - Mechanism skills: `flydocs-cloud` (Linear API) and `flydocs-local` (file-based)

package/template/flydocs/README.md

@@ -1,7 +1,8 @@
-# FlyDocs Cloud
+# FlyDocs
 
-This project uses FlyDocs Cloud - a hybrid workflow where:
-- **Work items (specs)** live in Linear
+This project uses FlyDocs - a hybrid workflow where:
+
+- **Work items (specs)** live in your issue tracker (local files or cloud provider)
 - **Project understanding** lives locally in this directory
 - **Framework config, rules & templates** live in `.flydocs/`
 
@@ -12,6 +13,7 @@ This project uses FlyDocs Cloud - a hybrid workflow where:
 **This `flydocs/` directory is your project's evolving knowledge base.**
 
 It grows alongside your project:
+
 - Update `context/` as your vision, stack, and standards evolve
 - Add to `knowledge/` as you make decisions and learn things
 - Expand `patterns/` as you discover project-specific conventions
@@ -25,7 +27,7 @@ FlyDocs will **never overwrite** these files during updates - they're yours to m
 
 ```
 .flydocs/                  # FRAMEWORK (replaced on FlyDocs update)
-├── config.json            # Provider settings, workspace config
+├── config.json            # Tier, provider settings, workspace config
 ├── version                # For upgrade detection
 ├── rules/                 # Workflow rules
 ├── scripts/               # Automation scripts
@@ -34,9 +36,7 @@ FlyDocs will **never overwrite** these files during updates - they're yours to m
 
 flydocs/                   # PROJECT CONTENT (evolves with project)
 ├── context/               # Project fundamentals
-│   ├── overview.md        # Vision, goals, scope (update as it evolves)
-│   ├── stack.md           # Tech stack, architecture (update as it changes)
-│   ├── standards.md       # Code conventions (update as they mature)
+│   ├── project.md         # Vision, stack, scope (update as it evolves)
 │   └── patterns/          # Project-specific patterns (grow over time)
 │
 ├── design-system/         # Design tokens (if using /design-setup)
@@ -59,19 +59,11 @@ flydocs/                   # PROJECT CONTENT (evolves with project)
 └── hooks.json             # Cursor hooks
 
 .claude/                   # Claude Code integration
+├── CLAUDE.md              # Claude Code instructions
 ├── agents/                # Subagent definitions
 ├── skills/                # Workflow skills
+├── commands/              # Slash commands
 └── settings.json          # Claude hooks
-│   ├── pm-agent/RULE.md          # Planning triggers
-│   ├── implementation-agent/RULE.md  # Build triggers
-│   ├── qe-agent/RULE.md          # Testing triggers
-│   ├── linear-integration/RULE.md
-│   └── ...
-└── commands/              # Slash commands
-
-.claude/
-├── CLAUDE.md              # Claude Code instructions
-└── commands/              # Symlinks to .cursor/commands/
 ```
 
 **Note:** Content folder name is configurable via `paths.content` in `.flydocs/config.json`.
@@ -82,51 +74,38 @@ flydocs/                   # PROJECT CONTENT (evolves with project)
 
 ### LOCAL (This Directory)
 
-| Content | Location | Purpose |
-|---------|----------|---------|
-| Project overview | `context/overview.md` | Vision, goals, scope |
-| Tech stack | `context/stack.md` | Architecture, dependencies |
-| Code standards | `context/standards.md` | Conventions, patterns |
-| ADRs | `knowledge/decisions/` | Why we made choices |
-| Feature docs | `knowledge/features/` | How features work |
-| Notes | `knowledge/notes/` | Learnings, gotchas |
-| Product docs | `knowledge/product/` | Personas, user flows |
+| Content          | Location               | Purpose              |
+| ---------------- | ---------------------- | -------------------- |
+| Project overview | `context/project.md`   | Vision, goals, scope |
+| ADRs             | `knowledge/decisions/` | Why we made choices  |
+| Feature docs     | `knowledge/features/`  | How features work    |
+| Notes            | `knowledge/notes/`     | Learnings, gotchas   |
+| Product docs     | `knowledge/product/`   | Personas, user flows |
 
 **Why local?** Agent needs instant access, changes with code, developers need it in IDE.
 
-### LINEAR (Cloud)
+### ISSUE TRACKER (Local or Cloud)
 
-| Content | Linear Location | Purpose |
-|---------|-----------------|---------|
-| Specs | Issues | Features, bugs, chores, ideas |
-| Status | Workflow states | BACKLOG → DONE progression |
-| Priorities | Priority field | Urgent, High, Medium, Low |
-| Assignments | Assignee | Who's working on what |
-| Assets | Attachments | Figma links, screenshots |
-| Progress | Comments | Decision log, impl notes |
+| Content     | Purpose                            |
+| ----------- | ---------------------------------- |
+| Specs       | Features, bugs, chores, ideas      |
+| Status      | BACKLOG → DONE progression         |
+| Priorities  | Urgent, High, Medium, Low          |
+| Assignments | Who's working on what              |
+| Progress    | Decision log, implementation notes |
 
-**Why cloud?** Team collaboration, stakeholder visibility, AI agent integration.
-
----
-
-## No Longer Exists
-
-These are **replaced by Linear**:
-
-- ~~`flydocs/specs/`~~ → Linear issues
-- ~~`flydocs/INDEX.md`~~ → Linear issue list
-- ~~`flydocs/ACTIVE.md`~~ → Linear "In Progress" view
-- ~~`flydocs/specs/assets/`~~ → Linear attachments
+**Local tier:** Issues stored as files in `flydocs/issues/`.
+**Cloud tier:** Issues stored in your provider (e.g., Linear). Enables team collaboration, stakeholder visibility, and AI agent integration.
 
 ---
 
 ## Quick Commands
 
 ```
-/start-session    - Check Linear status, plan work
-/capture          - Create new Linear issue
-/new-project      - Create project with product label/icon
-/implement        - Pick up issue from Linear
+/start-session    - Check issue status, plan work
+/capture          - Create new issue
+/new-project      - Create project with product label/icon (cloud tier)
+/implement        - Pick up issue for implementation
 /validate         - Test implementation
 /close            - Move issue to Done
 /flydocs-update   - Sync rules from source repo
@@ -144,10 +123,12 @@ If you have a Figma design system with tokens, run `/design-setup` to enable enh
 ```
 
 This creates:
+
 - `flydocs/design-system/token-mapping.md` - Your Figma → code translations
 - `flydocs/design-system/component-patterns.md` - Your reusable components
 
 And updates `.flydocs/config.json` with:
+
 ```json
 {
   "designSystem": {
@@ -168,15 +149,15 @@ And updates `.flydocs/config.json` with:
 ## Configuration
 
 See `.flydocs/config.json` for:
+
 - FlyDocs version
+- `tier` - "local" (free) or "cloud" (connected to provider)
 - `paths.content` - This folder name (default: "flydocs")
-- `provider.teamId` - Linear team ID
-- `workspace.activeProjects` - Array of active project IDs
+- `provider.type` - Issue tracker provider (null for local, "linear" for cloud)
+- `workspace.activeProjects` - Array of active project IDs (cloud tier)
 - `workspace.product` - Product identity (name, labelIds, icon, color)
 - `statusMapping` - Workflow state names
 
-See `.cursor/mcp.json` for MCP server configuration.
-
 ---
 
 ## More Information

package/template/flydocs/design-system/README.md

@@ -19,7 +19,8 @@ When using the `flydocs-figma` skill to extract designs, the AI needs to know:
 ### 1. Run /design-setup
 
 The `/design-setup` command will:
-- Connect to your Figma file via MCP
+
+- Connect to your Figma file via the Figma skill
 - Extract design tokens
 - Generate initial mappings
 - Create component pattern documentation
@@ -31,8 +32,8 @@ Edit `token-mapping.md` to define how Figma tokens translate to your code:
 ```markdown
 ## Colors
 
-| Figma Token | CSS Variable | Tailwind Class |
-|-------------|--------------|----------------|
+| Figma Token | CSS Variable        | Tailwind Class |
+| ----------- | ------------------- | -------------- |
 | Primary/500 | --color-primary-500 | bg-primary-500 |
 | Neutral/100 | --color-neutral-100 | bg-neutral-100 |
 ```
@@ -41,22 +42,26 @@ Edit `token-mapping.md` to define how Figma tokens translate to your code:
 
 Edit `component-patterns.md` to map Figma components to code:
 
-```markdown
+````markdown
 ## Button
 
 **Figma Component:** `Components/Button/Primary`
 
 **Code Pattern:**
+
 ```tsx
 <Button variant="primary" size="md">
   {children}
 </Button>
 ```
+````
 
 **Variants:**
+
 - Primary → variant="primary"
 - Secondary → variant="secondary"
 - Ghost → variant="ghost"
+
 ```
 
 ---
@@ -64,15 +69,17 @@ Edit `component-patterns.md` to map Figma components to code:
 ## File Structure
 
 ```
+
 flydocs/design-system/
-├── README.md              # This file
-├── token-mapping.md       # Figma token → code mapping
-├── component-patterns.md  # Figma component → code mapping
-└── tokens/                # Optional: extracted token JSON
-    ├── colors.json
-    ├── spacing.json
-    └── typography.json
-```
+├── README.md # This file
+├── token-mapping.md # Figma token → code mapping
+├── component-patterns.md # Figma component → code mapping
+└── tokens/ # Optional: extracted token JSON
+├── colors.json
+├── spacing.json
+└── typography.json
+
+````
 
 ---
 
@@ -115,12 +122,13 @@ Add version notes to track design system evolution:
 
 ### v1.0 (2023-06)
 - Initial design system
-```
+````
 
 ---
 
 ## Templates
 
 See `.flydocs/templates/design-system/` for:
+
 - `token-mapping.md` - Template for token documentation
 - `component-patterns.md` - Template for component documentation

package/template/manifest.json

@@ -1,5 +1,5 @@
 {
-  "version": "0.5.0-beta.1",
+  "version": "0.5.0-beta.10",
   "description": "FlyDocs Core - Manifest of all managed files",
   "repository": "github.com/plastrlab/flydocs-core",
 
@@ -18,9 +18,7 @@
         ".claude/skills/flydocs-figma",
         ".claude/skills/flydocs-estimates",
         ".claude/skills/flydocs-context-graph",
-        ".claude/skills/implementation-flow",
-        ".claude/skills/review-workflow",
-        ".claude/skills/spec-templates",
+        ".claude/skills/flydocs-context7",
         ".flydocs/scripts",
         ".cursor/agents"
       ]
@@ -51,7 +49,8 @@
         ".cursor/hooks.json",
         ".flydocs/version",
         ".flydocs/CHANGELOG.md",
-        "AGENTS.md"
+        "AGENTS.md",
+        ".env.example"
       ]
     },
 
@@ -60,7 +59,6 @@
       "paths": [
         ".env",
         ".env.local",
-        ".cursor/mcp.json",
         "flydocs/context/project.md",
         "flydocs/knowledge/**",
         "flydocs/design-system/**"
@@ -80,7 +78,6 @@
             "issueLabels",
             "statusMapping",
             "detectedStack",
-            "mcp",
             "skills",
             "designSystem",
             "aiLabor"
@@ -92,7 +89,6 @@
     "template_files": {
       "description": "Initial scaffolding - created on setup, then maintained by user as project evolves",
       "paths": [
-        ".env.example",
         "flydocs/context/project.md",
         "flydocs/knowledge/INDEX.md",
         "flydocs/knowledge/README.md",
@@ -151,13 +147,11 @@
           "flydocs-figma": ["SKILL.md", "references/"],
           "flydocs-estimates": ["SKILL.md", "references/"],
           "flydocs-context-graph": ["SKILL.md", "schema.md", "scripts/"],
-          "implementation-flow": ["SKILL.md"],
-          "review-workflow": ["SKILL.md"],
-          "spec-templates": ["SKILL.md"]
+          "flydocs-context7": ["SKILL.md", "cursor-rule.mdc", "scripts/"]
         }
       },
       ".cursor": {
-        "root": ["hooks.json", "mcp.json"],
+        "root": ["hooks.json"],
         "agents": [
           "pm-agent.md",
           "implementation-agent.md",
@@ -168,12 +162,7 @@
       ".flydocs": {
         "root": ["config.json", "version", "CHANGELOG.md"],
         "scripts": ["generate_manifest.py", "skill_manager.py"],
-        "hooks": [
-          "auto-approve.py",
-          "prefer-scripts.py",
-          "post-edit.py",
-          "prompt-submit.py"
-        ],
+        "hooks": ["auto-approve.py", "post-edit.py", "prompt-submit.py"],
         "templates": ["bug.md", "chore.md", "feature.md", "idea.md"]
       },
       "flydocs": {

package/template/.cursor/mcp.json

@@ -1,16 +0,0 @@
-{
-  "$schema": "https://cursor.sh/schemas/mcp.json",
-  "mcpServers": {
-    "linear": {
-      "command": "npx",
-      "args": ["-y", "@linear/mcp-server"],
-      "env": {
-        "LINEAR_API_KEY": "${LINEAR_API_KEY}"
-      }
-    },
-    "context7": {
-      "command": "npx",
-      "args": ["-y", "@anthropics/context7-mcp"]
-    }
-  }
-}