Agent Tools Discovery

agent_tool_discover.md

[SPEC] Agent Tool and Capability Discovery

Goal

• To systematically discover and document all features, tools, MCP servers, sub-agent support, skills (agentskills.io), and commands that the current agent environment reports to the LLM.
• Generate structured reports for each discovered category in the discovered_artifacts directory.

Context

• Current State: The agent has access to a variety of tools (standard and MCP-based), but there is no centralized documentation of these capabilities within the project for reference.
• Constraints: Must follow SDD v2.0 protocol. Reports must be generated without violating "Reporting Constraints" (except as requested for this specific discovery task).
• Files:
  • prompts/plans/agent_tool_discovery/agent_tool_discover.md (This spec)
  • prompts/plans/agent_tool_discovery/memory/ (Discovery logs)
  • prompts/plans/agent_tool_discovery/discovered_artifacts/ (Output)

Standards Compliance

• Table Architecture: N/A

Impact Surface

• Dependencies: None. This is a discovery and documentation task.
• Consumers: Developers/Users wanting to understand the full capability of the AI agent in this environment.

Risk & Rollback

• Potential Risks: None identified.
• Rollback Strategy: Delete the prompts/plans/agent_tool_discovery/ directory.

Phase 1: Tool & Command Discovery

Tasks

• Task 1.1: Enumerate all available tools provided by the IDE and MCP servers.
• Task 1.2: Identify built-in commands or "skills" (if any) supported by the agent.
• Task 1.3: Document the parameters and expected outputs for each tool.

Verification

• Check prompts/plans/agent_tool_discovery/discovered_artifacts/tools_list.md for completeness.

[WAIT FOR APPROVAL]

Phase 2: Advanced Capabilities & Agent Ecosystem

Tasks

• Task 2.1: Identify all connected MCP servers and their specific namespaces/tools.
• Task 2.2: Discover any sub-agent or multi-agent orchestration capabilities.
• Task 2.3: Check for integration with agentskills.io or similar skill definition frameworks.
• Task 2.4: Perform an exhaustive check for:
  • Context-sharing mechanisms (e.g., cross-agent memory).
  • Long-term memory or session persistence features.
  • Specialized model configurations (e.g., custom system prompts, temperature settings).
  • IDE-specific hooks or event listeners.
  • External API access beyond standard tools.
  • Plugin/extension support for the agent itself.

Verification

• Check prompts/plans/agent_tool_discovery/discovered_artifacts/mcp_infrastructure.md and prompts/plans/agent_tool_discovery/discovered_artifacts/advanced_capabilities.md.

[WAIT FOR APPROVAL]

Phase 3: Final Report Consolidation

Tasks

• Task 3.1: Generate a master summary of all capabilities.
• Task 3.2: Verify that all artifacts are correctly placed and formatted.

Verification

• Final review of discovered_artifacts/ content.

Final Deliverables

• prompts/plans/agent_tool_discovery/discovered_artifacts/tools_list.md
• prompts/plans/agent_tool_discovery/discovered_artifacts/mcp_infrastructure.md
• prompts/plans/agent_tool_discovery/discovered_artifacts/skills_and_commands.md
• prompts/plans/agent_tool_discovery/discovered_artifacts/advanced_capabilities.md
• prompts/plans/agent_tool_discovery/discovered_artifacts/master_summary.md

Discovery & Learning Log

• Record any hidden or undocumented tools found during the process.