Loading ToolBench
MCP server for 1C:Enterprise code analysis and AI assistance using the 1C:Naparnik API. Provides tools for syntax checking, code review, documentation search, and general 1C development questions.
Three tools with reasonable descriptions (50-150 chars each) and clear use cases. Input schemas are present with type definitions and descriptions. However, output schemas are completely undocumented, LLMs cannot predict what fields to expect from responses. Error handling is absent: no guidance on retryability, recovery steps, or actionable error messages. Tool naming follows verb_noun convention (ask_, explain_, check_) but lacks specificity about what 'ask' returns vs 'explain'. No parameter validation constraints (enums, ranges, patterns) visible in schemas. Session management via create_new_session parameter is stateful and non-idempotent, violating composition patterns.
Ask a free-form question to the 1C:Naparnik AI assistant. Use this tool for any 1C-related question that does not fit the specialised tools: general 1C development advice, architecture questions, explanations of concepts, syntax clarification, or anything else about the 1C:Enterprise ecosystem. Unlike all other tools (which always start a fresh session), this tool reuses the most recent session by default to preserve conversation context for follow-up questions. Set create_new_session=True to start from scratch when switching topics.
Check 1C:Enterprise code for syntax errors, logical issues, and performance problems. Use this tool for a technical correctness check: will the code compile, are there bugs, are there N+1 queries or other performance anti-patterns. In direct mode, syntax checking uses an upstream syntax-checker for precise validation, then adds logic and performance analysis via prompt. Automatically falls back to prompt-only mode if the direct tool is unavailable. For style/standards compliance, use review_1c_code instead. For an AI-proposed rewrite, use rewrite_1c_code instead.
Explain a specific 1C:Enterprise syntax element or language construct. Use this for focused questions about a particular keyword, operator, statement, or built-in function. Provide optional context to get a more relevant explanation. For broader questions, use ask_1c_ai instead.
Output schemas completely undocumented. No tool returns a documented structure, LLMs cannot predict response fields, forcing them to parse free-text strings and guess at downstream tool chaining.
No error handling guidance. Tools provide no recovery hints, retryability classification, or actionable error messages. A failed API call returns raw errors with no 'what to do next' guidance.
Stateful session management via create_new_session parameter violates idempotency. Repeated calls with same input produce different results (reuses vs creates session), breaking agent retry logic and causing duplicate side effects.
Inferred effective spec: 2026-07-28+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 60 | 2026-07-28+ | v2 |
No input validation constraints. check_1c_code accepts arbitrary code strings with no length limits, timeout guards, or injection prevention. Malicious or oversized inputs could hang the service.
Parameter descriptions lack format/constraint details. 'syntax_element' has no guidance on valid 1C keywords, length limits, or case sensitivity. LLMs will hallucinate invalid values.