Inspect AI system responses to detect whether prompt injection payloads bypassed safety restrictions, helping researchers evaluate defense effectiveness.
分析目标AI系统的响应，检测是否成功绕过安全限制
| Name | Required | Description | Default |
|---|---|---|---|
| response | Yes | 目标系统的响应内容 | |
| payload_id | No | 使用的载荷ID（可选） |
Changes observed during successful MCP inspections.
v1.0.0Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It says the tool 'detects whether bypass succeeded' but does not describe the return format (boolean, score, structured report), any confidence/threshold behavior, permission needs, or whether the analysis is deterministic. That is a significant gap for a tool whose entire value is its verdict output.
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?
A single front-loaded sentence with no filler. It is efficiently sized for the definition given, though the brevity borders on under-specification rather than optimal conciseness.
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?
With no annotations and no output schema, the description should explain what the analysis returns (verdict type, fields, interpretation), but it does not. An agent cannot tell what calling this tool yields beyond a vague 'detection'.
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 coverage is 100%, with both parameters (response, optional payload_id) fully documented in the schema. The description adds no additional meaning about parameter format or the role of payload_id, so the baseline 3 applies.
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?
States a specific verb (分析/analyze) and resource (目标AI系统的响应/target AI system's response), plus the detection goal (whether safety restrictions were bypassed). This clearly distinguishes it from list/get/search/generate siblings, though it doesn't explicitly differentiate from generate_report, which could also involve analysis.
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?
No guidance on when to invoke this tool, what prerequisite steps (e.g., run_test_sequence) should precede it, or how it relates to generate_report. Usage is only inferable from the name and workflow context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.