get_critical_warnings

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden of behavioral disclosure. It is exceptionally transparent: it reveals silent snapshot writes ('.vmcp/' or fallback), the exclusion of sim mode from snapshotting, log tailing behavior (tail_n), and the fallback scan for non-standard error keywords like TclStackFree/segfault/FATAL. It even discloses the limitation of runme.log for simulation runs and points to the actual xsim log location.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Although lengthy, the description is densely structured with numbered modes and bolded key terms, front-loading the core purpose and then systematically detailing each mode and parameter. Every sentence adds operational value—no filler. The format uses Markdown headers and lists to improve scannability for an AI agent.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The description is comprehensive for a complex diagnostic tool: it covers the main flow, edge cases (errors=0 but STATUS contains ERROR), mode-specific behavior (sim vs impl), snapshotting and comparison, and parameter details. An output schema exists, and the description also clarifies return content (Chinese diagnostic report with labels and repair suggestions). Even without review of the output schema, the description provides sufficient context to select and invoke the tool correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, so the description must compensate. It richly explains run_name (including mode routing), compare_with_last (diff report), and tail_n (range and per-sub-log behavior). However, session_id is only described as '目标会话 ID' (target session ID), which adds little beyond the parameter name itself, leaving a minor gap.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a specific verb+resource+scope: '提取并分类 CRITICAL WARNING / ERROR / 非标错误' (extract and classify critical warnings/errors/non-standard errors) and positions itself as a '统一失败诊断入口' (unified failure diagnosis entry). It clearly distinguishes itself from sibling monitoring/diagnostic tools like get_run_progress and get_timing_report by focusing on warning/error classification with actionable repair advice.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

Provides explicit usage guidance by describing three automatic diagnostic modes based on run_name (synth_*/impl_* vs sim_*). It explains when the tool is needed, e.g., '解决 messageDb 显示干净但 run 实际崩了' (solves the blind spot where messageDb looks clean but the run actually crashed), and clarifies that sim_* goes to a different log location. It also mentions compare_with_last behavior, giving clear context on when to use the diff functionality.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.