stride（stride 搜索）

STRIDE Threat Model Dispatcher

Dispatch parallel subagents covering all 6 STRIDE threat categories. Each category runs as an independent subagent analyzing the scoped code for that class of threat. STRIDE maps each category to a violated security property: Spoofing (Authentication), Tampering (Integrity), Repudiation (Non-repudiation), Information Disclosure (Confidentiality), Denial of Service (Availability), Elevation of Privilege (Authorization).

Supported Flags

Read ../../shared/schemas/flags.md for the full flag specification. This dispatcher supports all cross-cutting flags.

Framework Reference

Read ../../shared/frameworks/stride.md for the full STRIDE framework specification including threat descriptions, per-element applicability matrix, risk assessment guidance, and documentation templates.

Pre-flight Relevance Check

All 6 STRIDE categories are typically relevant to any codebase with user-facing functionality. However, perform a lightweight scan to confirm the codebase has code to analyze, and to build a targeted file list for each subagent.

For each category, use Glob and Grep to build a focused file list of the most relevant files. Pass this scoped list to the subagent rather than the full scope, so each subagent focuses on its area of expertise.

If --only is specified, dispatch only the listed categories.

Dispatch Category Subagents

CRITICAL: All Task tool calls MUST appear in the SAME response message. This is what triggers parallel execution. If you emit them across separate messages, they run sequentially and waste time.

Dispatch Table

Subagent Prompt Template

Each subagent Task call must include a FULLY self-contained prompt. Subagents get their own isolated context window and cannot see the main conversation.

Each subagent prompt must contain:

1. The concrete file list to analyze (resolved from scope, filtered for relevance).
2. The absolute path to the category SKILL.md to read and follow.
3. The flags to apply (--scope, --depth, --severity, --format, etc.).
4. The findings schema path (shared/schemas/findings.md) for output format.
5. An instruction to return findings only -- no summary, no cross-category commentary. The dispatcher handles consolidation.

Analyze the following files for STRIDE {LETTER} ({CATEGORY_NAME}) threats:

FILES:
{FILE_LIST}

STEP 1: Read the skill definition at:
{ABSOLUTE_PATH_TO_PLUGIN}/skills/{SKILL_NAME}/SKILL.md

STEP 2: Follow the workflow defined in that skill to analyze the listed files.
Focus on threats to the {SECURITY_PROPERTY} security property.

STEP 3: Read the findings schema at:
{ABSOLUTE_PATH_TO_PLUGIN}/shared/schemas/findings.md

STEP 4: Output findings in the schema format. Set metadata.framework to "stride"
and metadata.category to "{LETTER}".

FLAGS: --scope {SCOPE} --depth {DEPTH} --severity {SEVERITY}

IMPORTANT: Return ONLY the findings list. Do NOT produce a summary or
cross-category analysis. The dispatcher handles consolidation.

Launching

Emit one Task tool call per relevant category, ALL in a single response:

• subagent_type: "general-purpose"
• description: "STRIDE {LETTER} - {CATEGORY_NAME}"
• prompt: The fully self-contained prompt above, filled in for this category.

Do NOT emit Task calls one at a time. Do NOT wait between dispatches.

Consolidation

After ALL subagents return their results:

1. Merge Findings

Collect all findings from all subagent responses into a single list.

2. Deduplicate

Two findings are duplicates if they share the same location.file AND location.line (or overlapping line ranges). When duplicates exist:

• Keep the finding with the higher severity.
• Merge STRIDE category tags (a finding tagged T and I keeps both in references.stride).
• Note the duplicate in the retained finding's description.

3. Cross-Reference

For each finding, populate cross-framework references where known:

• references.owasp: Map to OWASP Top 10 category.
• references.cwe: Map to CWE identifier.
• references.mitre_attck: Map to ATT&CK technique ID.
• references.sans_cwe25: Map to SANS/CWE Top 25 rank if applicable.

Use the cross-framework mapping tables in stride.md for guidance.

4. Per-Element Threat Matrix

Build a summary matrix showing which STRIDE categories produced findings for each component or trust boundary. This mirrors the STRIDE per-element analysis approach:

5. Rank by Severity

Sort findings: critical > high > medium > low. Within the same severity, sort by confidence (high > medium > low).

6. Produce Aggregate Output

Wrap the consolidated findings in the aggregate output format from shared/schemas/findings.md, including categories_checked, categories_skipped, total_findings, and by_severity.

7. Present Results

Output the consolidated report in the requested --format. Include:

• Summary: categories checked, total findings, severity breakdown.
• Per-element threat matrix.
• Findings list in severity order.
• Trust boundary analysis: findings that cross trust boundaries are highlighted as higher risk.

Expert Mode

If --depth expert is set, after consolidation, launch red team subagents to simulate exploitation of the findings. Each red team agent receives the consolidated findings and constructs multi-step attack chains that cross STRIDE categories (e.g., Spoofing leads to Elevation of Privilege).

Read ../../shared/frameworks/dread.md for DREAD scoring criteria. Each finding receives a DREAD score in expert mode.

Launch red team agents as parallel Task calls (same single-response rule). Red team output is appended to findings with prefix RT and metadata.tool set to "red-team".