redteam-report-template — скилл для Claude Code: отчёт по ред-тиму
Скилл redteam-report-template задаёт структуру клиентского отчёта по ред-тиму: шесть разделов на находку и документ целиком, от executive summary до IoC и заявления об очистке.
MIT ★ 16 193 обновлён 2026-10-29
/plugin marketplace add elementalsouls/Claude-BugHunter
/plugin install claude-bughunter@elementalsoulsЧто делает redteam-report-template
Скилл задаёт вид клиентской находки: Subject, Observations, Description, Impact, Recommendation и PoC. В Observations только факты, Impact привязывает последствия к бизнесу клиента, а PoC идёт нумерованными шагами с точными запросами и payload.
Документ тоже собирается по фиксированному порядку: executive summary, детали работ, таблица рисков, находки от Critical к младшим, приложение с поверхностью атаки, IoC, заявление об очистке и аппендиксы. Severity описана бизнес-определениями с примерным диапазоном CVSS, а поле статуса помечает находку, которую клиент закрыл в середине теста. Отдельный блок разбирает типичные промахи: не отзывать находку, которая перестала воспроизводиться, не размывать Impact словами «потенциально может», не тащить recon-заметки в список находок.
Скилл про отчёты по договору с подписанным SOW. Для платформ баг-баунти он отсылает к report-writing и bugcrowd-reporting, потому что аудитория и тон там другие. Чтобы получить DOCX, нужен pandoc в системе (brew install pandoc) — без него конвертация не пойдёт, а число вставленных картинок проверяется через python-docx.
Какие задачи решает
- Свести находки в один клиентский отчёт по договору
- Описать Impact в терминах бизнеса, а не CIA-триады
- Отделить recon-заметки от подтверждённых находок
- Пометить находку, которую клиент закрыл в середине теста
- Собрать DOCX со встроенными скриншотами через pandoc
Примеры запросов агенту
- Собери финальный отчёт по этой находке в формате
redteam-report-template: Impact привяжи к бизнесу клиента, PoC распиши пошагово. - Проверь мои находки на соответствие структуре Subject / Observations / Description / Impact / Recommendation / PoC.
- Собери
REPORT_FINAL.docxизREPORT_FINAL.mdсо встроенными скриншотами.
Как установить redteam-report-template
1Через маркетплейс плагинов
/plugin marketplace add elementalsouls/Claude-BugHunter/plugin install claude-bughunter@elementalsoulsКоманда ставит плагин claude-bughunter целиком, со всеми его скиллами
2Через Skills CLI (Vercel)
npx skills add elementalsouls/Claude-BugHunter --skill redteam-report-template -a claude-code3Вручную: положить файлы
~/.claude/skills/redteam-report-template/SKILL.mdПапка скилла целиком; для одного проекта — .claude/skills/redteam-report-template/
1Через Skills CLI (Vercel)
npx skills add elementalsouls/Claude-BugHunter --skill redteam-report-template -a codex2Вручную: положить файлы
.agents/skills/redteam-report-template/SKILL.mdДля всех проектов — ~/.agents/skills/
1Через Skills CLI (Vercel)
npx skills add elementalsouls/Claude-BugHunter --skill redteam-report-template -a cursor2Вручную: положить файлы
.cursor/skills/redteam-report-template/SKILL.mdCursor читает и .agents/skills/, и .claude/skills/
1GitHub CLI
gh skill install elementalsouls/Claude-BugHunter skills/redteam-report-template2Вручную: положить файлы
.github/skills/redteam-report-template/SKILL.md1Через Skills CLI (Vercel)
npx skills add elementalsouls/Claude-BugHunter --skill redteam-report-template -a gemini-cli2Командой платформы
gemini skills install https://github.com/elementalsouls/Claude-BugHunter --path skills/redteam-report-templateВручную: положить файлы
.devin/skills/redteam-report-template/SKILL.mdСтарый путь .windsurf/skills/ тоже читается
Claude.ai
Упакуйте папку в zip и загрузите в Settings → Capabilities → Skills
По шагам, с командами для каждой платформы: Как установить скилл в Claude Code: пять способов
Вопросы о redteam-report-template
Как установить redteam-report-template в Claude Code?
Через маркетплейс плагинов: сначала /plugin marketplace add elementalsouls/Claude-BugHunter, затем /plugin install claude-bughunter@elementalsouls.
Что делает redteam-report-template?
Задаёт формат клиентского отчёта по ред-тиму: шесть разделов на каждую находку, таблица severity, статусы, порядок разделов документа и сборка в DOCX.
Чем он отличается от отчётов для баг-баунти?
Он рассчитан на внешний ред-тим по договору, где отчёт читают CISO и IT-Sec. Для H1, Bugcrowd и Intigriti скилл советует report-writing и bugcrowd-reporting.
Исходный текст
SKILL.md · MIT
--- name: redteam-report-template description: Client-facing red-team deliverable format — codifies the Subject / Observations / Description / Impact / Recommendation / PoC structure used for external red-team engagements (not bug-bounty platform reports). Different audience, different tone, different cadence. Built from an authorized engagement deliverable where 14 findings were packaged into a 52KB MD + 2.2MB DOCX with 16 embedded screenshots. Use when the engagement is "external red team for an enterprise client" (not H1/Bugcrowd/Intigriti), when generating the final report, when the client has specified a custom report format, or when packaging findings into DOCX/PDF. sources: authorized-engagement-deliverable, enterprise-redteam-report-conventions report_count: 1 --- ## When to use Use this skill for **client-deliverable** reports: - External red-team engagements with a signed SOW - Pentest reports going to a CISO / IT-Sec team (not a triager) - Findings that will be reviewed by both technical and non-technical stakeholders - Reports that need DOCX/PDF output (not just markdown / platform UI) Do NOT use for: - Bug-bounty platform submissions (use `report-writing` / `bugcrowd-reporting` instead) - Quick proof-of-concept memos - Internal team writeups --- ## The 6-section format per finding This is the canonical structure each finding follows: ```markdown ## Finding F##: <descriptive title> **Severity:** Critical / High / Medium / Low / Informational **Status:** Confirmed / Patched mid-engagement / Suspected (1 signal) **CVSS 3.1:** <score> (<vector>) **Affected Asset:** <URL / IP / app name> ### 1. Subject <One-line statement of the issue. Plain English, no jargon.> ### 2. Observations <Bulleted list of what was observed during testing. Concrete facts only — no interpretation yet.> - <Observation 1> - <Observation 2> - ... ### 3. Description <Technical explanation of the vulnerability. 2-4 paragraphs. Reader should understand WHY the observations indicate a vulnerability, what the underlying flaw is.> ### 4. Impact <What an attacker could achieve. Concrete attacker outcomes, NOT generic CIA triad statements. Tie to the client's business — money, data, reputation, regulatory exposure.> ### 5. Recommendation <Specific, actionable remediation. Vendor patch, configuration change, code-level fix. Avoid "implement security best practices" — say what specifically.> ### 6. Proof of Concept (PoC) <Steps to reproduce, numbered. Include the exact HTTP requests, payloads, tools used.> **Step 1:** <action> ```http <full HTTP request or curl one-liner> ``` **Step 2:** <action> ``` <response excerpt> ``` **Screenshot:**  ``` --- ## Severity & status disciplines ### Severity table (client-facing — different from CVSS-only) | Severity | Business definition | CVSS rough range | |---|---|---| | Critical | Direct revenue/data loss without prerequisites | 9.0-10.0 | | High | Full account/system takeover with limited prerequisites | 7.0-8.9 | | Medium | Significant data exposure or partial compromise | 4.0-6.9 | | Low | Information disclosure with limited exploitation path | 0.1-3.9 | | Informational | Hygiene finding, no immediate exploit | N/A | ### Status field (red-team-specific) This is the field that distinguishes red-team deliverables from bug-bounty reports. Use one of: - **Confirmed** — reproduced multiple times, with full PoC - **Confirmed; patched mid-engagement** — was reproducible, client patched during the test window (still ship the finding — see `mid-engagement-ir-detection`) - **Confirmed; partially reproducible** — works but needs specific conditions - **Suspected (1 signal)** — single indicator, not confirmed (rare — usually drop) - **Out-of-band** — finding from passive recon, not actively tested --- ## Mistakes to avoid (from authorized-engagement) ### 1. Don't retract findings that stopped reproducing If a finding was confirmed and then stopped working, that is almost always a CLIENT PATCH, not a finding-was-false. The correct response is "Confirmed; patched mid-engagement" with timestamps showing when it broke. See `mid-engagement-ir-detection`. ### 2. Don't hedge in the Impact section Bad: "An attacker could potentially be able to access user data, which may lead to..." Good: "An attacker reads any user's profile data. Demonstrated on test user `[email protected]` at 14:22 IST." ### 3. Don't generic-CIA the impact Bad: "Loss of confidentiality and integrity of customer data" Good: "Read access to 247,000 customer records including PAN cards, addresses, GST numbers. India DPDPA Section 33 mandates 72-hour breach disclosure to DPB." ### 4. Don't list every recon finding as a finding Recon notes (subdomains found, ports open, technologies fingerprinted) belong in a separate **Recon / Attack Surface** appendix, not the findings list. A finding must have an attacker-attainable outcome. ### 5. Don't bury the PoC Each finding MUST have reproducible steps. The PoC section is what proves the finding to a skeptical reader. If you can't write the PoC clearly, the finding probably isn't ready to ship. --- ## Document-level structure ``` 1. Executive Summary (1 page, non-technical) - Engagement overview (dates, scope) - Risk posture summary (heat-map: <X critical, Y high, Z medium...>) - Top 3 strategic recommendations - Comparison to industry baseline (optional) 2. Engagement Details - Scope (in-scope, out-of-scope, exclusions) - Methodology (recon → exploit → reporting; or align with PTES / OSSTMM) - Tools used - Timeline (start / end / key milestones) - Team 3. Risk Summary Table | F# | Title | Severity | Status | |----|-------|----------|--------| | F01 | ... | Critical | Confirmed | ... 4. Findings (one per ## section, in severity order — Critical first) 5. Attack Surface / Recon Appendix - Subdomains discovered - Open ports / services - Technology fingerprints - APKs found - Credentials in breach corpora (count + sample only — redact) - Identity-fabric map (IdP, MFA posture) 6. Indicators of Compromise (IoCs) - Source IPs used during testing (so SOC can correlate) - User-Agent strings - Test accounts created - Files uploaded (with cleanup status) 7. Cleanup Statement - Confirmation that all test artifacts (accounts, uploads, persistence) were removed - Outstanding cleanup items requiring client action 8. Appendices (raw output, screenshots index, full target list) ``` --- ## DOCX generation pipeline (markdown → docx with embedded images) ```bash # Prerequisite: pandoc installed brew install pandoc # Convert pandoc REPORT_FINAL.md \ -o REPORT_FINAL.docx \ --resource-path=engagement_log/poc \ --reference-doc=~/.claude/skills/redteam-report-template/templates/reference.docx \ --toc \ --toc-depth=2 \ --highlight-style=tango # Verify image count python3 -c " from docx import Document d = Document('REPORT_FINAL.docx') imgs = [r for r in d.part.rels.values() if 'image' in r.target_ref] print(f'Embedded images: {len(imgs)}') print(f'Paragraphs: {len(d.paragraphs)}') print(f'Headings: {sum(1 for p in d.paragraphs if p.style.name.startswith(\"Heading\"))}') " ``` ### Image filename convention ``` screenshots/F<NN>_<descriptive>.png Examples: F01_locked_accounts.png F02a_saml_landing.png F02b_saml_ca_block_page.png F03_sqli_timing_chart.png F15_saml_metadata.png ``` Variants get letter suffixes (F02a, F02b). Always zero-pad finding number. --- ## Writing tone — for client deliverables | Section | Tone | |---|---| | Subject | Plain English, jargon-free, 1 line | | Observations | Bulleted facts, past tense ("observed that...") | | Description | Technical but accessible; assume CISO reader | | Impact | Business-translated; tie to revenue/regulation | | Recommendation | Imperative, specific, actionable | | PoC | Operator-level technical; copy-pasteable | **Always:** - Use past tense for observations ("The endpoint returned a 200 status code") - Use present tense for descriptions of the flaw ("The application does not validate...") - Use imperative for recommendations ("Apply patch ... by ...") - Number reproduction steps; never "first... then... also..." **Never:** - "Could potentially" — prove it or drop it - "It might be possible" — same - "We recommend implementing security best practices" — say which one specifically - "The application is vulnerable to..." without saying what specifically --- ## Audience translation — same finding, different framing Example: hardcoded JWT in APK | Section | Technical framing | CISO framing | Board framing | |---|---|---|---| | Impact | "JWT signing key extracted from APK enables forging admin tokens" | "Anyone with the customer-facing mobile app can read any customer's invoice" | "A leaked secret in our mobile app lets attackers impersonate users" | The same finding's Impact paragraph should cover both ends — start with the business outcome, then drop into technical detail. --- ## Findings that are sometimes wrongly excluded Red-team deliverables should include — not just bug-bounty payable bugs: - **Information disclosure that helps attack mapping** (CodeIgniter debug toolbar leaking routes, version banners on appliances) — Medium - **User enumeration** (Microsoft OneDrive 302 vs 404 differential) — Medium - **Pre-existing security state observations** (247 accounts already locked by external attacker; weak password policies) — Informational with stakeholder relevance - **Defensive observations** (SOC patched our SQLi within 30 min — evidence of working detection) — Informational/positive - **Sister-app pattern issues** (same code template across 7 apps) — Medium (multiplied blast radius) Bug bounty would reject most of these. Red-team deliverables embrace them — the client paid for the assessment to know. --- ## Mid-engagement events to document Beyond findings themselves, the deliverable should include: - Detected SOC responses (timestamps when defenses kicked in) - Concurrent external attacker activity (if any was observed) - Findings the client patched during the engagement (with PoC pre-patch as evidence) - Tooling failures (e.g., MCP timeout, CAPTCHA not solvable) — these affect what was/wasn't testable Each gives the client context about their real-world detection capability, which often matters more than the findings themselves. --- ## Template library (where to put canned text) Maintain reusable boilerplate in: ``` ~/.claude/skills/redteam-report-template/templates/ executive_summary.md # Reusable exec summary skeleton methodology.md # Standard methodology section cleanup_statement.md # Standard cleanup language reference.docx # Pandoc style template (fonts, headings, colors) cover.docx # Cover page template ``` Don't write these from scratch each engagement; clone and customize. --- ## Quality checks before delivery Pre-delivery checklist: - [ ] Every finding has all 6 sections populated (no "TBD") - [ ] Every finding has at least one screenshot or HTTP-level evidence - [ ] Every PoC includes redactions for client PII (mask emails, phone numbers, IDs) - [ ] Every screenshot is referenced in the MD with a relative path that resolves - [ ] DOCX render check — image count matches MD reference count - [ ] Severity ordering: Critical findings first, then High, etc. - [ ] Executive summary is updated to match final findings (count, themes) - [ ] Cleanup statement explicitly says what was created and what was removed - [ ] IoC section enables the SOC to reconstruct what they saw - [ ] Spell-check (especially client company name, product names) - [ ] All tool versions noted in methodology - [ ] Status field set correctly on every finding (especially patched-mid-engagement) --- ## Bridge to neighboring skills - `report-writing` — bug-bounty platform reports (different format, different audience) - `redteam-mindset` — informs what counts as a finding worth shipping - `mid-engagement-ir-detection` — informs the "patched mid-engagement" status pattern - `evidence-hygiene` — informs screenshot redaction discipline - `m365-entra-attack`, `enterprise-vpn-attack`, etc. — each provides finding-templates specific to its attack surface --- ## Anti-patterns - **DO NOT** write a 50-page report for 3 findings — pad-by-page erodes credibility - **DO NOT** use the same severity for everything — calibrate - **DO NOT** copy-paste OWASP top-10 boilerplate into Description sections - **DO NOT** include findings without PoCs — they read as speculative - **DO NOT** skip the Recommendation section's specificity — "patch and review" doesn't help - **DO NOT** mix bug-bounty CVSS scoring with red-team severity unthinkingly — context differs (e.g., a Medium on a CVSS basis can be Critical for the client if it touches a regulated dataset) --- ## Real engagement metric (authorized-engagement) For calibration: - 14 findings shipped (2 Critical, 4 High, 5 Medium, 3 Low/Info) - 18 screenshots embedded - 52,737 bytes markdown / 2,262,484 bytes DOCX - 414 paragraphs, 123 headings in DOCX - 16 inline images (2 were inline in MD-only edge cases) - Time-to-deliverable: ~6 hours after engagement close for first draft These numbers are typical for a 1-week external red-team engagement on a mid-size enterprise. Scale down for short tests, up for full purple-team exercises. --- ## Related Skills & Chains - **`triage-validation`** — This template ingests findings that have ALREADY passed the 7-Question Gate. Engagement flow: every finding through `triage-validation` first → only validated findings → `redteam-report-template` packaging. Skipping triage produces a deliverable padded with informational noise that erodes client trust. - **`evidence-hygiene`** — The DOCX with 16 embedded screenshots only works if evidence was captured systematically throughout the engagement. Engagement flow: `evidence-hygiene` discipline at session start → timestamped, organized screenshot folder → `redteam-report-template` consumes that folder to populate Evidence blocks. - **`redteam-mindset`** — The Subject / Observations / Description / Impact / Recommendation / PoC structure assumes the operator already thinks like a red-teamer (impact-first, blast-radius framing). Engagement flow: `redteam-mindset` loaded at engagement start → findings collected with red-team framing baked in → `redteam-report-template` produces deliverable without rewriting every Impact section. - **`mid-engagement-ir-detection`** — Defensive-action findings (SOC patches mid-test, new IPS rules deployed, account lockouts triggered by external attacker) are first-class findings in red-team deliverables. Engagement flow: `mid-engagement-ir-detection` captures behavior-change events → each becomes its own Subject in the deliverable, framed as "client capability observation" not as "bug we missed." - **`report-writing`** + **`bugcrowd-reporting`** — Bug-bounty platform reports use DIFFERENT structure (one finding per submission, platform-specific severity scoring, OOS-clause counters). Engagement flow: if engagement mode is `bug-bounty` per project memory → use `report-writing` / `bugcrowd-reporting` instead. This template is ONLY for external red-team / enterprise client deliverables.
© Sachin Sharma · MIT · elementalsouls/Claude-BugHunter