/security-flow-deterministic-gates

name: security-flow-deterministic-gates
description: "Phase 3 Deterministic Gates of security-flow"
disable-model-invocation: true
user-invocable: false
alwaysApply: false
tags: ["security", "workflow-phase"]
baseSchema: docs/schemas/phase.md

<security_flow_deterministic_gates>

<description_and_purpose> Run approved deterministic high-severity gates before lifecycle AI analysis. </description_and_purpose>

<workflow_context> Phase 3 for development, change, PR, and pipeline reviews. </workflow_context>

<phase_steps> 1. Select deterministic gates 2. Run bounded commands 3. Preserve sanitized results 4. Choose lifecycle branch </phase_steps>

<execute_gates step="3.1" subagent="executor" role="Bounded deterministic security-gate operator" subagent_required_model="Claude Haiku 4.5, GPT-5.6 Terra, Gemini 3.7 Flash, GPT-5.6 Luna">

1. USE SKILL `subagent-directives`. 2. USE SKILL `security`. 3. Run approved installed local tools only. 4. Preserve command, version, configuration, timestamps, exit status, and sanitized output reference. 5. Do not interpret beyond deterministic rule/severity results. 6. Return unchanged source finding records. 7. Update `security-flow-state.md`.

</execute_gates>

<branch step="3.2"> Return exactly one token with the records:

• Any deterministic high+: HIGH+.
• No deterministic high+: CLEAN.
• Tool error or incomplete evidence: ERROR; stop and report the tool error; never call the gate clean.

</branch>

<validation_checklist>

• Every required gate ran successfully.
• High+ records remain unchanged.
• Clean means zero high+ and no tool failure.
• Preserved evidence and output references contain no secret values.

</validation_checklist>

</security_flow_deterministic_gates>