MCP

MCP Documentation

Reference for the Damper MCP server. Connect an AI assistant to the same workspace that runs demand, roadmap, releases, and internal notes — then use tools for tasks, context, and progress.

Setup

Add the Damper MCP server to your AI assistant. The server uses streamable HTTP transport — no local installation required.

Recommended Workflow

AI agents should follow this structured workflow for best results. The workflow ensures agents have full context before coding and properly track their progress.

Project Context

Store and retrieve project documentation that helps AI agents understand your codebase. Context is organized into named sections (e.g., overview, conventions, testing) and can be resolved from the files a change is likely to touch.

Task Management

Core tools for working with tasks. Agents pick tasks, lock them to prevent conflicts, track progress, and mark work complete.

Setup

Add the Damper MCP server to your AI assistant. The server uses streamable HTTP transport — no local installation required.

{
  "mcpServers": {
    "damper": {
      "type": "http",
      "url": "https://api.usedamper.com/mcp",
      "headers": {
        "Authorization": "Bearer dmp_..."
      }
    }
  }
}

Authentication

All MCP requests are authenticated via your API key in the Authorization header.

Authorization: Bearer dmp_your_api_key

Recommended Workflow

AI agents should follow this structured workflow for best results. The workflow ensures agents have full context before coding and properly track their progress.

Call list_projects if you do not already know the Damper project ID for this repo or task.
Call get_project_context for the project index and task-relevant critical rules. For feature work, call the read-only prepare_work with likely files and inspect per-path fixtureCoverage; use resolve_context for smaller work. Verify fallback suggestions against precise context and code.
Call list_tasks to see available tasks sorted by priority.
Call start_task to lock a task and receive its full spec and project context.
Work on the task. Use add_note and add_commit to track progress.
When finished, load the relevant checklist branches, then call complete_task with itemId-based confirmations and evidence for done items, or abandon_task with a handoff summary.

Task Management

Core tools for working with tasks. Agents pick tasks, lock them to prevent conflicts, track progress, and mark work complete.

Tool	Description
`list_tasks`	List tasks with optional filtering by status, type, or label. Use query to search titles, descriptions, implementation plans, user stories, or exact labels before paginating. Returns tasks sorted by priority (weighted demand score) with available dates such as created, updated, and due dates.
`get_task`	Get full task details including description, implementation plan, subtasks, notes, and linked feedback.
`create_task`	Create a new task with title, description, type, and optional implementation plan.
`update_task`	Update task fields like title, description, status, type, labels, or effort estimate.
`delete_task`	Delete a planned task that has no commits. Use with care; this is permanent.
`start_task`	Lock a task for the current agent. Returns the task spec, project context, and critical rules. Prevents other agents from working on it.
`add_note`	Add a progress note to a task. Use for decisions, observations, or status updates. Notes are visible to future agents.
`add_commit`	Record a git commit against a task. Pass the commit hash and message. Helps track what code was written.
`complete_task`	Mark a task as complete and release the lock. Requires itemId-based checklist confirmations with structured proof for done items and explicit checklistScope dimensions when appliesTo, when, or riskDomains rules are used.
`abandon_task`	Release a task lock without completing it. Requires a handoff summary describing what was done and what remains.

Subtasks

Break large tasks into smaller pieces. Agents check off subtasks as they complete each part, giving visibility into progress.

Tool	Description
`create_subtask`	Add a checklist subtask to a task. Provide a title.
`update_subtask`	Mark a subtask as done or not done.
`delete_subtask`	Remove a subtask from a task.

Task Review

Request a review of completed work before finalizing.

Tool	Description
`review_task`	Submit a task for review with a summary of changes. Returns review feedback or approval.

Project Context

Tool	Description
`get_project_context`	Load the project context index. Pass taskId when known. The criticalRuleMode option defaults to relevant; use all for an intentional full audit or none for an index-only response. Call this first in every session.
`resolve_context`	Resolve likely touched paths to context and checklist scope. Defaults to fallbackMode=minimal and detailMode=compact; broad fallback and full detail are available for compatibility or deliberate exploration. Pass caller-known changeKinds when possible. Fallback metadata and warnings guide exploration and are not governing truth until verified against precise context and code.
`prepare_work`	Build a deterministic, read-only packet with compact per-path fixtureCoverage. A represented status has scope route_representation, validationStatus not_checked, and scenarioCoverage not_inferred: representedBy and unrepresentedStructuredRoutes show route representation only, not validated assertions or every scenario under a broad route. Every gap needs a disposition; critical and important gaps are requiredBeforeCompletion, while advisory gaps may use documented justification.
`audit_context_health`	Audit taxonomy validity, route-to-fixture coverage, failing or weak fixtures, missing baselines, stale knowledge, overlaps, and fallback usage.
`list_context_routes / upsert_context_route / compare_context_resolution`	Manage validated structured path routes. Structured matches take precedence while Markdown routing remains a migration fallback.
`list_context_fixtures / upsert_context_fixture / validate_context_routing`	Store reviewed expectations derived from intended project truth and code, never copied resolver output. Fix unverified fallback routing before adding a fixture, and run validate_context_routing before relying on representation. Add a precise route and fixture for materially distinct risks, contracts, verification, or cross-surface behavior; a broad route's fixture is not scenario proof.
`list_context_bundles / upsert_context_bundle`	Manage coherent groups of sections, example paths, and verification profiles. Every update creates an immutable version snapshot.
`report_context_feedback`	Report missing, irrelevant, or incorrectly routed knowledge using privacy-safe metadata without prompts or source code.
`list_context_sections`	List all context sections with their titles, tags, and appliesTo targets.
`get_context_section`	Get the full content of a specific context section by name. Supports hierarchical paths (e.g., api/architecture) and wildcards (api/*).
`get_section_blocks`	Get the heading structure of a large context section. Returns block titles for selective loading.
`get_section_block_content`	Load a specific block from a context section by heading. Useful for token-efficient reading of large docs.
`update_context_section`	Create or update a context section. Provide content, optional tags, appliesTo targets, and critical rules.
`delete_context_section`	Remove a context section by name.
`sync_project_context`	Sync project context from a local CLAUDE.md or similar file to Damper. Useful for initial setup.

Feedback

Access user feedback and weighted votes to inform task priorities. Link feedback items to tasks to track what customer requests led to shipped features.

Tool	Description
`list_feedback`	List feedback items with optional filtering by type, status, or search query. Returns items with vote counts, demand scores, and available created, updated, first-seen, and last-seen dates.
`get_feedback`	Get full feedback details including description, votes, voter tiers, and linked tasks.
`update_feedback`	Update a feedback item's status, type, or internal notes.
`link_feedback_to_task`	Link a feedback item to a task. When the task ships, feedback submitters can be notified.

Code Templates

Maintain code consistency with reusable templates for services, components, tests, and other patterns. Templates teach AI agents your project's conventions.

Tool	Description
`list_templates`	List all code templates with their names, tags, and descriptions.
`get_template`	Get the full content of a code template by name.
`update_template`	Create or update a code template with content, tags, and description.
`sync_templates`	Sync code templates from local files to Damper. Reads template files and uploads them.

Module Registry

Track monorepo structure with package paths, ports, dependencies, and descriptions. Helps AI agents understand which package to modify.

Tool	Description
`list_modules`	List all registered modules with their paths, ports, and descriptions.
`get_module`	Get full details of a module including dependencies and configuration.
`update_module`	Create or update a module entry with path, port, description, and dependencies.
`sync_modules`	Sync module registry from package.json files in the monorepo.

Changelogs

Create and manage changelog entries. When AI agents complete public tasks, draft changelog entries are created automatically.

Tool	Description
`list_changelogs`	List changelog entries with optional filtering by status (draft, published).
`create_changelog`	Create a new changelog entry with title, content, version, and linked tasks.
`update_changelog`	Update an existing changelog entry's title, content, version, or status.
`add_to_changelog`	Add completed tasks to an existing changelog entry; private tasks are shown publicly only as other updates.
`publish_changelog`	Publish a draft changelog entry. Sends notifications to subscribers of linked roadmap items.
`delete_changelog`	Delete a changelog entry.

Settings & Utilities

Configure project settings, get agent-specific instructions, and access utility tools.

Tool	Description
`get_project_settings`	Get current project settings including the completion checklist index, standards refs, scope guidance, default labels, and notification preferences.
`update_project_settings`	Update project settings like the structured completion checklist, standards refs, default task labels, or notification rules.
`get_agent_instructions`	Get agent-specific instructions that should be followed during task execution. Includes critical rules and workflow guidance.
`report_issue`	Report a bug or issue with the Damper platform. Useful for agents encountering unexpected behavior.
`open_public_page`	Get the URL for a project's public page (roadmap, changelog, or feedback board).