LIVE
US Military Avoids Close Call After AI-Generated Intelligence Hallucination18/09/26|Study finds gender bias in GPT models is not reduced but reshaped across generations17/09/26 · OpenAI|Architectural tweaks may break conventional scaling law exponents16/09/26|OpenAI publishes a framework for reporting model misalignment16/09/26 · OpenAI|NVIDIA's Vera Rubin NVL72 Debuts in MLPerf Inference v6.116/09/26 · NVIDIA|OpenAI expands ChatGPT advertising with Sponsored Agents16/09/26 · OpenAI|OpenAI moves into advertising with 'Sponsored Agents'16/09/26 · OpenAI|Google DeepMind Introduces Gemini 3.8 Live and Its Extended Thinking Variant15/09/26 · Google DeepMind|What's at stake in AI's trillion-dollar infrastructure bet15/09/26|A Flaw in Chain-of-Thought Safety Monitoring14/09/26|Stellar Colosseum: A Multi-Agent System for Long-Horizon Mathematical Research14/09/26|Apple Code Hints Siri Could Be Swapped for ChatGPT or Claude14/09/26 · Apple|US Military Avoids Close Call After AI-Generated Intelligence Hallucination18/09/26|Study finds gender bias in GPT models is not reduced but reshaped across generations17/09/26 · OpenAI|Architectural tweaks may break conventional scaling law exponents16/09/26|OpenAI publishes a framework for reporting model misalignment16/09/26 · OpenAI|NVIDIA's Vera Rubin NVL72 Debuts in MLPerf Inference v6.116/09/26 · NVIDIA|OpenAI expands ChatGPT advertising with Sponsored Agents16/09/26 · OpenAI|OpenAI moves into advertising with 'Sponsored Agents'16/09/26 · OpenAI|Google DeepMind Introduces Gemini 3.8 Live and Its Extended Thinking Variant15/09/26 · Google DeepMind|What's at stake in AI's trillion-dollar infrastructure bet15/09/26|A Flaw in Chain-of-Thought Safety Monitoring14/09/26|Stellar Colosseum: A Multi-Agent System for Long-Horizon Mathematical Research14/09/26|Apple Code Hints Siri Could Be Swapped for ChatGPT or Claude14/09/26 · Apple|
SkillNewofficial publisher

Veille CVE ciblée (skill nAIvigate)

Skill : A reusable method: instructions, procedures, sometimes scripts or templates · requires claude-code

Filters vulnerability bulletins against your actual estate and sorts them into 72 h, 30 days, to plan.

Goes from 47 bulletins read to the 4 that concern you, with the source of every line.

Overall score
8.4 /10
Security: nothing to flag
Utility
8.0
Reliability
7.0
Security
No caveat on record
Ease
1 click · Comfortable

Security reflects where data goes, requested permissions and incidents on record at the verification date. It is not a guarantee of zero risk: check the deployment mode and the connections you enable.

What it does, concretely

Vulnerability watch drowns in volume: every week, dozens of CERT, NVD and vendor bulletins of which a handful actually affect the organisation. This skill takes a list of product + version pairs (never machines or addresses) and a bulletin source, pasted or searched over the period, then keeps only the matches.

Each retained vulnerability is qualified: CVSS score, known exploitation (KEV catalogue, vendor notice), available fix, your exposure. Ranking crosses exploitation and exposure rather than score alone, giving three blocks: act within 72 h, within 30 days, plan. What is discarded is counted, never silently dropped.

The bulletin ends with a 'what this bulletin does not know' section: unconfirmed versions, vendors with no source found. It helps prioritise; the patching decision and testing stay with the team.

What it offers

  • ·Input: product + version + exposure inventory (internal, perimeter, SaaS), no hosts or addresses.
  • ·Accepted sources: pasted text, a period to search, or specific CVEs to qualify.
  • ·Per-line qualification: CVSS, known exploitation, fix, exposure, action, source.
  • ·Three urgency blocks crossing exploitation and exposure, plus the count of discarded bulletins.
  • ·'Version to confirm' flag whenever a match is not certain.

In practice

Price
Free
Commercial use
Commercial use: yes
Available on
claude, claude-code
Host
claude-code
Licence
MIT
FR interface
Yes

Security and data

Rating
Security: nothing to flag
Your data
Data stays local
Maintenance
Active

Links and repository checked on 19 September 2026 · added on 19 September 2026

How to install Veille CVE ciblée (skill nAIvigate)

1. Download https://naivigate.com/downloads/skills/veille-cve-ciblee.skill (direct link, no sign-up). 2. Claude.ai: open the file in a conversation, or Settings › Capabilities › Skills › Import. The skill triggers on its own when the request matches. 3. Claude Code: unzip the file and copy the folder into ~/.claude/skills/ (personal) or .claude/skills/ in the project (shared). 4. Other editors: SKILL.md is plain Markdown, paste it into the project instructions. 5. Before first use, read the skill's 'Data never to provide' section.

Official documentation

In the same category

Whole category Frame (methods, guardrails)