security-intel-brief
pramseier-tenb/security-intel-brief
A Claude skill that produces a leadership-ready security intelligence brief (PDF) for the vendors and software you choose, down to specific versions.
Instalação
npx convez-ai pramseier-tenb/security-intel-brief1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
README
A tradução para o português deste README ainda está na fila.
security-intel-brief
A Claude skill that produces a leadership-ready security intelligence brief (PDF) for the vendors and software you choose — down to specific versions.
For each target it collects, from live sources at run time:
- CVEs from the NVD API (exact-version matching via CPE)
- CISA KEV status (known exploited vulnerabilities — confirmed in-the-wild)
- Latest versions and end-of-life dates (endoflife.date)
- Vendor security advisories and recent security news (web search)
It then assigns a per-product risk rating (Critical/High/Medium/Low) with stated rationale and renders a polished PDF: executive summary, risk overview table, recommended actions, and per-product detail sections (version/patch status, CVE tables with KEV flags, advisories, news, sources).
Install
Claude (Cowork / Claude Code): download security-intel-brief.skill from this repo (or the latest release) and install it via your skills settings — or copy this folder into your skills directory.
Use
Say something like:
Security briefing on Cisco IOS XE 17.9.4a and Apache Tomcat 9.0.85 for Monday's leadership meeting
or just:
What vulnerabilities affect our Tomcat servers?
If you don't name targets, the skill shows an interactive form to pick vendors, products, versions, lookback window, and report title.
Contents
| Path | Purpose |
|---|---|
SKILL.md | The skill instructions (workflow, risk-rating guidance, accuracy rules) |
references/data-sources.md | API endpoints, query parameters, response fields, gotchas |
scripts/build_report.py | Renders the report JSON to a branded PDF (reportlab) |
assets/selection_form.html | Interactive target-selection form |
evals/evals.json | Test prompts used during development |
Notes
- All vulnerability data is fetched at run time — nothing is answered from model memory.
- The PDF builder needs
reportlab(pip install reportlab); the skill installs it automatically when missing. - NVD is rate-limited (~5 requests/30s without an API key); the skill spaces out calls.