AGENT SKILL

conops-template

Concept of Operations document creation — executive summary, threat actor profiling, attack narrative, kill chain design, communication plan, deconfliction.

License 许可
Apache-2.0
所属仓库
Akapi895/raptix
最近更新
2026-11-06
Stars
★ 0
Forks
0

01 前言 Front Matter

name
conops-template
metadata
{ "tags": "conops, kill-chain, threat-model, operation-design", "subdomain": "planning", "when_to_use": "create CONOPS, design operation, threat model, plan attack", "upstream_ref": "Soundwave CONOPS template — Concept of Operations document generator" }
description
Concept of Operations document creation — executive summary, threat actor profiling, attack narrative, kill chain design, communication plan, deconfliction.
allowed-tools
Read Write Edit

02 资源文件 1

03 SKILL.md 正文

Concept of Operations (CONOPS) Generator

The CONOPS bridges the legal RoE and the tactical OPPLAN. It must be readable by a CEO while containing enough detail for operators.

When to Use

  • After plan/roe.json exists
  • User says "create CONOPS", "design the operation", "build threat model"
  • Before OPPLAN can be generated

Prerequisites

Read plan/roe.json first — scope and boundaries constrain the CONOPS.

See ../references/schema-quick-reference.md for the CONOPS, ThreatActor, KillChainPhase, and DeconflictionPlan schema fields.

Workflow

Step 1: Interview the User

Budget: 2 questions max for CONOPS/Threat Profile (soundwave.md CRITICAL_RULES → Question Budget). The remaining fields below are DEFAULTED or AGENT-DRAFTED, not asked — the tier table and RoE Constraint→Profile Implication table in threat-profile/SKILL.md already give deterministic defaults for motivation and initial access once the tier is picked, so re-asking them wastes a turn.

Question 1 — Threat actor tier (single-select, use threat-profile skill for detailed profiling):

  • a) Opportunistic external attacker (low)
  • b) Targeted cybercriminal (medium)
  • c) APT / nation-state (high)
  • d) Insider threat
  • e) Custom — describe

Derive motivation and initial access vector from the picked tier via threat-profile/references/adversary-archetypes.md (default) — do not ask separately. Only deviate if the operator's free-text answer (via allow_other) already states a motivation/vector explicitly.

Question 2 — Success criteria — the crown-jewel / measurable win condition. Required; no default (every engagement needs an explicit end-state).

Agent-drafted, not asked:

  • Attack narrative — write the 2-3 sentence scenario yourself from RoE scope + tier + success criteria. This is the agent's job, not the operator's homework.
  • Ultimate objectives — derive from success criteria; do not ask as a separate dimension.
  • Communication plan — REMOVED. CONOPS.communication_plan is DEPRECATED (see SCHEMA_REFERENCE in soundwave.md); ContactPlan (contact-template skill) owns this now. Asking about it here duplicates a question and writes to a dead field.
  • Deconfliction method — default to standard red-team markers/headers (see deconfliction-template reference) unless the operator's RoE/contact answers already flagged a SOC integration endpoint; do not spend a dedicated question on it.

Step 2: Design Kill Chain

Based on RoE scope + threat profile, select applicable phases. See references/kill-chain-templates.md.

Key rule: Don't include phases outside RoE scope. Recon-only engagement → only recon phase.

Step 3: Generate Documents

  1. plan/conops.json — matching CONOPS schema
  2. plan/deconfliction.json — matching DeconflictionPlan schema

Step 4: Validate

  • Executive summary contains no jargon or tool names
  • Kill chain phases align with RoE scope
  • All MITRE ATT&CK technique IDs are valid
  • Timeline has concrete date ranges
  • At least 2 success criteria defined

Generation Rules

  1. Executive summary = non-technical — no tool names, no jargon
  2. Threat actor TTPs must reference MITRE ATT&CK IDs
  3. Kill chain scoped to RoE — no exploitation phase in recon-only engagement
  4. Timeline uses absolute dates — never relative
  5. Deconfliction defaults to standard red-team markers/headers unless a SOC integration endpoint was already flagged; communication_plan is DEPRECATED — write coordination details to ContactPlan instead

04 同仓库更多 Skills