ATLAS MCP Server

ATLAS (Adaptive Task & Logic Automation System) is a project, knowledge, and task management system designed for LLM Agents. It is implemented as a Model Context Protocol (MCP) server, allowing AI models to manage structured data through a 3-node architecture.

The 3-Node Model

• Project: The top-level container for metadata, status, completion requirements, and output formats.
• Task: Actionable steps belonging to a project. Tasks inherit project context and support priority, status, and dependency tracking.
• Knowledge: A searchable repository of information linked to projects or tasks, supporting domain categorization and citations.

System Integration

• Relationships: Projects contain tasks; both projects and tasks can be enriched with knowledge items.
• Dependencies: Both projects and tasks support dependency relationships for complex workflows.
• Search: A unified search capability allows finding projects, tasks, or knowledge via content, metadata, or relationships.

+-------------------------------------------+
                  |                PROJECT                    |
                  |-------------------------------------------|
                  | id: string                                |
                  | name: string                              |
                  | description: string                        |
                  | status: string                             |
                  | urls?: Array<{title: string, url: string}>| 
                  | completionRequirements: string            |
                  | outputFormat: string                      |
                  | taskType: string                          |
                  | createdAt: string                         |
                  | updatedAt: string                         |
                  +----------------+--------------------------+
                            |                    |
                            |                    v                    v
+----------------------------------+ +----------------------------------+
|               TASK               | |            KNOWLEDGE             |
|----------------------------------| |----------------------------------|
| id: string                       | | id: string                       |
| projectId: string                | | projectId: string                |
| title: string                    | | text: string                     |
| description: string              | | tags?: string[]                  |
| priority: string                 | | domain: string                   |
| status: string                   | | citations?: string[]             |
| assignedTo?: string              | | createdAt: string                 |
| urls?: Array<{title: string,     | |                                  |
|   url: string}>                  | | updatedAt: string                 |
| tags?: string[]                 | |                                  |
| completionRequirements: string   | |                                  |
| outputFormat: string             | |                                  |
| taskType: string                 | |                                  |
| createdAt: string                 | |                                  |
| updatedAt: string                 | |                                  |
+----------------------------------+ +----------------------------------+

You can use the atlas_deep_research tool to conduct comprehensive research tasks. This tool is designed to gather information from various sources, including government programs, private foundations, and community organizations.

When prompting for deep research, you should specify:

• The specific scope of the research (e.g., grant opportunities for a specific region or project type).
• Priority criteria (e.g., application deadlines, eligibility requirements, funding amounts).
• Preferred tools for augmentation (e.g., using perplexity for search-augmented generation or firecrawl for web scraping).
• Instructions for database management (e.g., clearing the database before starting or keeping it up to date).
• The desired output location for the final report.

Example prompt structure:

# Research Task: [Topic]

Conduct research on [Subject] using atlas_deep_research. Focus on [Specific Details].

Research should prioritize:
- [Priority 1]
- [Priority 2]

Use the perplexity search tool for LLM augmented searching and firecrawl tools for web scraping. Keep the atlas db completely up to date. 

Start by clearing the db.

# Grant Research Task: Covington, KY Community Projects

Conduct comprehensive research on publicly available grants for solo-owned community projects in Covington, Kentucky using atlas_deep_research to gather information. Focus on funding opportunities specifically suitable for individual owners of community gardens and similar local initiatives. Identify grant programs from government sources (federal, state, and local), private foundations, and community organizations that support urban agriculture and neighborhood improvement projects.

Research should prioritize:

- Current open application periods and upcoming deadlines
- Grant amounts and funding cycles
- Eligibility requirements specific to individual applicants rather than organizations
- Any special programs targeting the Covington area or Northern Kentucky region
- Application processes and documentation requirements
- Past successful community garden projects in the area

Include information on matching grant opportunities and any technical assistance programs available to support the application process.

Perform extensive research and provide a detailed report on your findings in /Users/casey/Developer/github/atlas-mcp-agent/workspace.

Use the perplexity search tool for LLM augmented searching (search-augmented-generation); use the firecrawl tools for web scraping when necessary. Keep the atlas db completely up to date with all details.

Perform extensive research. Include specific details about the grants. Include details on how to apply for the grants.

Start by clearing the db.