Loading ToolBench
Authenticated local credential vault with explicit v1 migration, fixed-operation broker, MCP and loopback REST. Resolves {{PLACEHOLDER}} tokens before sending to LLM.
EnigmAgent MCP server has critical definition gaps. All 4 tools lack input parameter descriptions, the schema shows parameter names and types but no descriptions explaining what each parameter controls or accepts. Tool descriptions are present but generic (10-50 chars), below the 50-200 char baseline for LLM-optimized descriptions. No output schemas are documented. The tools appear to be vault/secret management operations, but descriptions do not explain WHEN to use each tool, what it returns, or how they compose. No error handling guidance visible. Tool naming follows verb_noun convention (vault_status, vault_list, resolve, resolve_text) but descriptions are too terse to guide LLM selection.
Resolve a {{PLACEHOLDER}} token to its secret value
Resolve all {{PLACEHOLDER}} tokens in a text string
List all secret keys in the vault
Check vault status and authentication state
All tools lack parameter descriptions. enigmagent_resolve has a 'placeholder' parameter with type string but no description explaining format (e.g., 'The placeholder token to resolve, formatted as {{KEY_NAME}}'). enigmagent_resolve_text has 'text' parameter with no description of expected content or token format.
No output schemas documented. Tools return data but LLMs cannot plan downstream calls or extract fields. vault_status and vault_list should document return structure (e.g., {status: string, authenticated: boolean} and {keys: string[], count: number}).
Tool descriptions are too generic (10-50 chars, below 50-200 char baseline). 'Check vault status and authentication state' and 'List all secret keys in the vault' lack context on WHEN to call each tool, what distinguishes them, or what the LLM should do with results.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 41 | 2026-07-28+ | v2 |
No error handling guidance. If resolve fails (placeholder not found, vault locked, authentication failed), LLM has no recovery path. Should document: 'If placeholder not found, call vault_list() to see available keys. If vault is locked, authentication may be required.'
Tool composition unclear. resolve and resolve_text appear to do similar things (resolve placeholders) but descriptions do not explain when to use each. resolve_text likely processes multiple tokens in a string; resolve processes one. This distinction must be explicit.