Agent2 exposes 4 tools with minimal documentation. All tools lack proper descriptions (under 20 chars or generic placeholders). Input schemas are present but parameters lack descriptions. The template tool ('example_tool') is a placeholder with no real implementation. Code-review tools (count_lines, detect_language, check_complexity) have trivial descriptions that do not explain WHEN to use them or what they return. No output schemas documented. No error handling guidance. No parameter constraints (enums, ranges). This is a template/demo server, not production-ready.
Estimate code complexity based on nesting and branching.
Count the number of lines in the code.
Detect the programming language from code content.
Replace with database, API, search, or file-processing logic.
Tool descriptions are under 20 characters or generic placeholders. 'Replace with database, API, search, or file-processing logic' is not a description, it's a TODO comment. LLMs cannot determine when to select these tools.
No parameter descriptions. The 'code' parameter in count_lines, detect_language, and check_complexity lacks any explanation of expected format, constraints, or examples. LLMs cannot infer what 'code' means (source code? bytecode? pseudocode?).
No output schemas documented. Tools return results but the response structure is not specified. LLMs cannot plan downstream calls or extract the right fields.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 42 | 2026-07-28+ | v2 |
No error handling guidance. Tools provide no recovery hints if they fail. An LLM receives a bare error and has no guidance on whether to retry, ask the user, or try a different tool.
example_tool is a placeholder with no real implementation. It signals this is a template, not a production server. Placeholder tools confuse LLMs and waste reasoning cycles.