CreditDelta MCP Guardian
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| GEMINI_MODEL | No | Optional Gemini model name (e.g., gemini-3.6-flash). Used when GEMINI_API_KEY is set. | |
| GEMINI_API_KEY | No | Optional API key for Google Gemini report generation. If not provided, deterministic template reports are used. | |
| SEC_USER_AGENT | Yes | User agent string identifying the client to SEC (e.g., 'Your Name your.email@example.com'). |
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": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| search_companyA | Find a public company and its SEC CIK using a name or ticker. |
| list_metric_periodsC | List recent SEC values for a supported financial metric. Supported metrics: revenue, cash, current_debt, net_income, operating_cash_flow. |
| compare_financialsC | Compare the two most recent SEC values for one financial metric. |
| evaluate_tool_callC | Ask MCP Guardian whether a proposed tool call is safe to execute. |
| guarded_actionC | Evaluate and safely execute, block, or queue a protected local action. |
| resolve_approvalB | Approve or reject one pending guarded action. |
| list_pending_actionsA | List actions waiting for explicit user approval. |
| list_audit_eventsB | Return recent redacted Guardian decisions and execution outcomes. |
| run_security_benchmarkA | Measure Guardian correctness and decision latency on adversarial cases. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| analyze_company | Create instructions for analyzing a company's financial change. |
| security_review | Create instructions for reviewing an action with MCP Guardian. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| supported_metrics | Return the financial metrics understood by this starter server. |
TDQS
Scored across 9 tools
The tools fall into two distinct clusters (guardian/security and financial data) with clear boundaries. Within the guardian cluster, guarded_action and evaluate_tool_call differ by scope (local actions vs. tool calls), and list/resolve/benchmark tools are unambiguous. Financial tools are clearly distinct by resource (company, metrics, comparison). No two tools could be easily confused.
All tool names follow a consistent verb_noun pattern in snake_case (guarded_action, resolve_approval, list_pending_actions, etc.). Each starts with a clear verb and uses a descriptive noun, making the pattern predictable across both clusters.
With 9 tools, the server is well-scoped for its combined purpose of security guard and financial data lookup. Each tool serves a necessary function without redundancy, and the count is comfortably within the ideal 3–15 range.
The guardian workflow is fully covered: evaluation, execution, approval, pending list, audit log, and benchmarking. Financial data supports search, metric listing, and comparison, but lacks a tool to retrieve raw multi-period values or detailed financial statements, which agents may need to infer from comparisons.