dev-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| DEV_MCP_ALLOWED_DIRS | No | Colon-separated list of allowed directories for file access. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
| prompts | {
"listChanged": true
} |
| resources | {
"subscribe": false,
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| read_fileB | Read the contents of a file with optional line-range slicing and line numbering. |
| write_fileB | Write content to a file with atomic commit or append mode. |
| edit_fileB | Replace target string with replacement text in a file with occurrence control and diff preview. |
| batch_replaceB | Search and replace a string across multiple files matching a glob pattern. |
| file_infoB | Get metadata, line count, permissions, and size for a file or directory. |
| delete_pathB | Delete a file or an empty directory. |
| delete_recursiveB | Recursively delete a file or directory tree. Use with caution. |
| list_directoryC | List files and directories inside a given directory with size and clean filtering. |
| glob_filesB | Find files matching a glob pattern (e.g. '/*.ts', 'src//*.py') with automatic noise filtering. |
| create_directoryA | Create one or more directories (like mkdir -p). |
| search_codeC | Search for a text or regex pattern across codebase with ripgrep, noise exclusion, and context lines. |
| run_commandB | Run a shell command and return its output. Always set a reasonable timeout. Avoid destructive commands (rm -rf, etc.). |
| git_statusB | Show the working tree status (git status). |
| git_diffB | Show staged and unstaged changes (git diff). |
| git_logA | Show recent commit history (git log). |
| git_branchC | List local branches (git branch). |
| git_addB | Stage file(s) for commit (git add). |
| git_commitC | Commit staged changes with a message (git commit). |
| git_create_branchB | Create and switch to a new git branch (git checkout -b). |
| install_git_hooksA | Install automated pre-commit hook to sync Project Brain and context with native terminal git commits. |
| init_contextC | Initialize or auto-detect project context. Scans project files to detect tech stack, languages, and frameworks. |
| get_contextB | Get the full project context as formatted text. Optionally filter by dot path (e.g. 'tech_stack.languages'). |
| update_contextB | Update a specific field in the project context using dot notation (e.g. 'project.description', 'conventions.code_style'). |
| add_to_context_listB | Add an item to a list field in the context (e.g. tech_stack.languages, key_files, active_tasks, goals). |
| add_key_fileC | Register a key file with its purpose in the project context. |
| add_taskC | Add an active task to the project context. |
| add_decisionC | Log an architecture decision record (ADR) in the project context. |
| export_context_markdownC | Export the project context as a Markdown document for sharing with other LLMs and AI tools. |
| remove_context_itemC | Remove an item from a list in the project context by index. |
| reset_contextB | Reset the project context to default values (clears all data). |
| project_infoC | Get information about the current project environment. |
| desktop_statusA | Check desktop accessibility status, permissions, and supported surfaces. |
| desktop_list_appsA | List all running GUI desktop applications. |
| desktop_list_windowsC | List all open windows with titles, dimensions, and window IDs. |
| desktop_snapshotC | Capture the accessibility tree of an application with actionable @refs. |
| desktop_findC | Search for UI elements matching a role, name, or text in an app. |
| desktop_clickB | Click an element using its qualified snapshot reference (@snapshot_id:eN). |
| desktop_typeC | Insert text into a text field target using its qualified ref. |
| desktop_pressB | Send a keyboard shortcut or key combo (e.g. 'ctrl+c', 'return', 'escape', 'tab'). |
| desktop_scrollC | Scroll an element or page by direction (up, down, left, right). |
| desktop_launchC | Launch a desktop application and optionally wait for its window to appear. |
| desktop_close_appC | Close an application gracefully or force terminate it. |
| desktop_focus_windowC | Bring an application or window to the foreground and focus it. |
| desktop_clipboardB | Read, write, or clear system clipboard ('get', 'set', or 'clear'). |
| desktop_verify_elementC | Verify if a GUI element exists and assert its state (e.g. enabled, selected). |
| desktop_screenshotC | Take a screenshot of an application window and save to target path. |
| desktop_actionC | Universal dispatcher executing ANY agent-desktop CLI command (hover, drag, wait, is, get, toggle, batch, etc.). |
| get_prdA | Retrieve a PRD document (PRD, VISION, SCOPE, REQUIREMENTS). Returns the content of the specified document. |
| update_prdC | Create or update a PRD document (PRD, VISION, SCOPE, REQUIREMENTS). |
| summarize_prdC | Summarize key requirements from PRD documents into a concise bullet list. |
| compare_work_to_prdC | Compare current project context and tasks against PRD requirements to check alignment. |
| get_current_phaseB | Get the current active phase and its status. |
| set_phaseA | Set the current active development phase. Only tasks in this phase are considered 'in scope'. |
| complete_phaseC | Mark the current phase as completed and optionally advance to the next phase. |
| next_phaseC | Mark current phase as complete and advance to the next phase in sequence. |
| list_phasesB | List all phases and their status (completed, active, pending). |
| add_milestoneB | Define a milestone with a target date and associated features. |
| list_milestonesC | Show all milestones and their status. |
| create_taskC | Create a new task with phase, priority, dependencies, and description. |
| update_taskC | Update a task's status, priority, or description by ID. |
| complete_taskC | Mark a task as completed with optional notes. |
| next_taskA | Get the highest-priority task that is ready to work on (dependencies met, current phase). |
| blocked_tasksB | List all tasks that are blocked, either by dependencies or explicitly. |
| list_tasksC | List all active tasks, optionally filtered by phase or status. |
| generate_handoffA | Generate a session summary capturing what was done, what's next, and current project state for handoff to another AI. |
| get_last_handoffA | Retrieve the most recent AI handoff document so a new AI session has full context. |
| list_handoffsB | List all AI session handoffs with dates. |
| generate_project_brainC | Generate a complete aggregated Project Brain dashboard showing current status, phase, active/blocked tasks, recent changes, and recommendations. |
| next_best_actionB | Intelligently calculate and recommend the next best task to work on based on phase, priority, dependencies, and file hints. |
| start_sessionA | Start a development session. Loads state, PRD, active tasks, and latest handoff to generate a detailed Session Brief for the AI. |
| end_sessionB | End the current session, update progress, and generate a handoff file (does not auto-commit changes). |
| prepare_commitB | Analyze modified workspace files and generate a suggested conventional git commit message and summary for approval. |
| commit_changesB | Stage all workspace changes and commit them to Git with the approved commit message. |
| validate_architectureC | Scan the codebase and validate imports/technologies against the project's architectural guidelines and forbidden tools. |
| build_dependency_graphC | Scan files and construct a complete mapping of dependencies and references in the project. |
| impact_analysisC | Calculate direct and transitive downstream files affected if a given file is modified. |
| sync_workspaceB | Scan the physical workspace (Git diff, new files, current branch, commits) and compare to the Todo plan to auto-detect finished work. |
| verify_workC | Perform deep sanity, test, git diff, and architecture checks against a task to output a reality verification dossier. |
| project_dashboardB | Generate a unified Project Manager Dashboard combining phase state, todo progress, git status, architecture violations, and a project health score. |
| extract_requirementsC | Parse PRD markdown files to extract structured requirements and save them into requirements.json. |
| verify_requirementsC | Scan codebase to map requirement keywords and verify their implementation status and feature coverage. |
| verify_outcomeC | Perform outcome audits verifying user flows, business goals, and objective completion based on target tasks. |
| assess_project_readinessB | Scan the workspace and evaluate its onboarding readiness score (0-100%) against required Project Brain documents. |
| project_readiness_reportC | Generate a formatted Markdown project readiness report showing missing files, scoring audit, and recommended actions. |
| bootstrap_projectB | Auto-generate template files and directories (PRD, milestones, active tasks, config yaml) to instantly onboard an unknown repository. |
| adopt_existing_projectB | Scan an existing legacy codebase (directories, package lists, readmes) and generate context, PRDs, milestones, and audit tasks. |
| doctorB | Check the diagnostic health and configuration integrity of the Project Brain and MCP server. |
| reset_project_brainB | Wipe all Project Brain configuration, tasks, and context metadata inside the selected project. Leaves git and source code untouched. |
| archive_project_brainA | Archive current Project Brain settings and context files to .project_brain_archive/ before a reset or pivot, preserving history. |
| create_project_workspaceA | Create a new isolated project workspace directory, initialize git, bootstrap the project context/brain, and set up first milestone. |
| get_dev_mcp_profileA | Get the active dev-mcp tool profile and registered module count. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| project-overview | A complete project overview prompt that injects the full project context for AI assistants. |
| code-review | Code review guidelines specific to this project's conventions and tech stack. |
| architectural-decision | Template for recording an Architecture Decision Record (ADR). |
| feature-plan | Template for planning a new feature with tasks, branch strategy, and milestones. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| Project Context (JSON) | The full project context document as JSON. |
| Project Context (Markdown) | The full project context as a Markdown document for sharing. |
TDQS
Scored across 91 tools
Several clusters of tools are hard to tell apart: git_commit vs commit_changes vs prepare_commit, add_task vs create_task vs update_task, next_task vs next_best_action, generate_project_brain vs project_dashboard, and generate_handoff vs end_session. delete_path vs delete_recursive and desktop_action vs the specific desktop_* verbs also overlap, forcing agents to guess which variant to use. Descriptions help somewhat, but there are too many near-duplicates across domains.
Mostly consistent snake_case, but conventions vary by module: git_* and desktop_* are prefixed, while file operations (read_file, search_code, batch_replace) and brain tools (doctor, next_best_action, next_task) are not. Some names are vague or unprefixed in a way that makes module boundaries unclear, though the overall style remains readable.
91 tools covering file I/O, git, desktop GUI automation, PRD management, phases, tasks, handoffs, architecture analysis, and project bootstrapping is far beyond what can be coherently scoped as one server. Individual capabilities may earn their place, but the aggregate is extreme and should be split into multiple servers.
The surface is effectively over-complete for development workflows, covering file editing, search, git flows, PRD/phase/task lifecycle, handoffs, and verification. Minor gaps exist (e.g. git push/pull/tag/stash, richer git log filtering) but run_command can compensate.