ai-secure-by-design

Audit evidence plan

sbd_evidence_plan

Read-onlyIdempotent

Generates per-control evidence and audit plans for AI Security Wiki checks, showing what the engine can verify, what it found, and the test reviewers need. Reports observations, not verdicts.

Instructions

For each AI Security Wiki control: what the engine can check, what it found here, and the evidence and audit test a reviewer needs. Reports observations, never a verdict.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "sbd_evidence_plan"
  }
}

Input Schema

Name	Required	Description	Default
`tier`	No
`manifestPath`	No	Manifest path relative to the workspace root. Default privacy-security-manifest.yaml.

Schema Changelog

Changes observed during successful MCP inspections.

First observedv0.1.0Nov 5, 2026

TDQS

B3/5.0

Behavior3/5

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

Annotations already establish readOnly, idempotent, non-destructive, closed-world behavior, so the safety profile is covered. The description adds one genuinely useful behavioral fact beyond that: it reports observations and 'never a verdict,' which sets the authority boundary of the output. It does not say how the workspace is inspected or what the report structure looks like, so the addition is real but narrow.

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

Conciseness4/5

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

Two tight sentences with no filler, and the scope disclaimer is placed at the end where it reads as a caveat. Minor cost: the opening is a prepositional fragment rather than a leading verb, so the agent must read the whole sentence to learn what the tool produces.

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

Completeness3/5

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

With no output schema, the description carries the burden of explaining returns, and it does partially enumerate the report contents (checks, findings, evidence, audit test). It still leaves the output format, the effect of 'tier,' and the manifest-reading behavior unstated, which is a meaningful gap for a two-parameter tool with a 50%-described schema.

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

Parameters2/5

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

Schema coverage is only 50%: manifestPath carries a description with its default, but the 'tier' enum (baseline/advanced) is undocumented anywhere. The description never mentions either parameter, so it does nothing to compensate for the unexplained tier distinction, which likely changes what the plan covers.

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

Purpose4/5

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

The description names the resource precisely: a per-control evidence plan covering 'what the engine can check, what it found here, and the evidence and audit test a reviewer needs.' That is noticeably different from siblings like sbd_check or sbd_explain_control, though the description never names those siblings or states the generating verb explicitly, leaving the agent to infer the differentiation.

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

Usage Guidelines2/5

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

There is no explicit when-to-use statement and no alternatives are named among the nine sibling tools. The phrase 'a reviewer needs' hints at an audit-preparation context, but that is inference rather than guidance, and nothing says when to prefer this over sbd_check or sbd_list_controls.

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

Deploy Server