Critical gaps across all dimensions. Tool descriptions are extremely brief (5-20 chars), far below the 34-char minimum baseline. No parameter descriptions visible for most tools. Input schemas present but minimal, many tools lack parameter type annotations. No output schemas documented. Error handling absent. Security concerns: 'open-the-door' and 'transfer-sui' are high-risk operations with no confirmation, dry-run, or permission gates. Naming is inconsistent (kebab-case with hyphens; 'get-sui-address' vs 'transfer-sui'). No tool follows the verb_noun pattern clearly. Overall, this reads as a template/demo server, not production-ready.
get coin market in coingecko
Get current coin price
get sui address
get sui balance
Open the door by the pass
send notice to telegram
transfer sui to one address
Descriptions critically short (5-20 chars). Rubric baseline is 34-392 chars; these are 1/3 of minimum. LLMs cannot determine when/why to select tools. Examples: 'send notice to telegram' (24 chars), 'get sui address' (14 chars), 'Open the door by the pass' (25 chars).
No parameter descriptions visible. 'pass' param in 'open-the-door' has no description. 'to' and 'amount' in 'transfer-sui' lack descriptions explaining format/constraints. Rubric baseline: 100% of A+ tools have param descriptions.
High-risk operations lack confirmation/dry-run. 'transfer-sui' (IRREVERSIBLE) and 'open-the-door' (WRITE) have no confirmation step, dry-run mode, or permission gates. Agents can execute destructive actions without safeguards.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 38 | <=2025-11-25 | v2 |
No output schemas documented. Tools return unstructured text responses (e.g., 'Current bitcoin price: {...}'). LLMs cannot parse results reliably or chain to downstream tools. Rubric baseline: 100% of A+ tools have documented return types.
No error handling or recovery guidance. Tools return raw errors (e.g., 'Failed to get bitcoin price: [error]'). LLMs cannot determine if error is retryable, user-fixable, or fatal. No actionable next steps provided.