mcp-AgentCore
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| AGENTCORE_URL | Yes | The base URL of the AgentCore server (e.g., http://localhost:3000). | |
| AGENTCORE_TOKEN | No | Optional JWT token for authenticating with the AgentCore server. Required only when authentication is enabled on the server. |
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 | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| investigate_incidentA | Run an AgentCore incident investigation: retrieves similar past incidents from episodic memory, plans tool calls with an LLM, queries logs/metrics/runbooks, and synthesizes a report. Waits for completion by default and returns the synthesized incident report. |
| get_investigationA | Fetch an investigation's status, checkpoint timeline (per-step status and durations), and final report if completed. |
| search_incident_memoryA | Semantic search over AgentCore's episodic memory of past incidents and their outcomes. Use to find how similar incidents were resolved before. |
| export_postmortemA | Export a completed investigation as a Markdown post-mortem (summary, probable cause, affected systems, next actions, checkpoint timeline). |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 4 tools
Each tool has a clearly distinct purpose: investigating, fetching investigation status/reports, exporting post-mortems, and searching past incident memory. There is no meaningful overlap or ambiguity between them.
All tool names follow the same verb_noun snake_case pattern: investigate_incident, get_investigation, export_postmortem, search_incident_memory. The naming is predictable and consistent.
Four tools is well-scoped for an incident investigation server. Each tool covers a distinct step in the workflow without bloat or redundancy.
The core investigation lifecycle is covered: run, retrieve, search memory, and export post-mortem. A listing or cancellation capability for investigations would improve completeness, but agents can still accomplish the primary workflow without dead ends.