Standalone MCP server exposing Signet signing, verification, and content hash tools for cryptographic action receipts
Three tools with clear cryptographic purposes and basic parameter documentation. All tools have descriptions (10-20 chars baseline met) and input schemas with typed parameters. However, descriptions are minimal (avg 85 chars vs 194 baseline), lack WHEN/WHY context, and output schemas are not documented. Parameters lack format constraints and validation guidance. No error handling documentation or recovery paths. Tools follow verb_noun naming correctly but lack composition guidance for multi-step workflows.
Generate SHA-256 hash of content for integrity verification
Sign content using Ed25519 private key to create a cryptographic receipt
Verify a signature against content using Ed25519 public key
Output schemas not documented. LLMs cannot infer what sign_content, verify_signature, and hash_content return. Missing return type documentation prevents downstream tool chaining and forces agents to guess field names.
Descriptions lack WHEN/WHY context. 'Sign content using Ed25519 private key to create a cryptographic receipt' states WHAT but not WHEN to use this vs alternatives or what prerequisites exist. Descriptions should be 50-200 chars with actionable context.
Parameter descriptions lack format constraints. 'Ed25519 private key in base64 format' and 'Ed25519 signature in base64 format' state the format but not length, encoding validation, or error handling if invalid. Should specify: 'Base64-encoded Ed25519 private key (88 chars, no padding)'.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 56 | 2026-07-28+ | v2 |
No error handling documentation. Tools do not document what happens if private key is invalid, signature verification fails, or base64 decoding errors. LLMs need recovery guidance: 'If verification fails, check that public key matches the signing key.'
Metadata parameter in sign_content is under-specified. Type is 'object' with no schema, no description of expected fields, no examples. LLMs cannot determine what metadata to include or how to structure it.