The Copilot appears in five places, each with its own prompt key, rate limit, and regulatory purpose:

Technical File suggestions (prompt key: suggest) – drafts content for each of the eight Annex VII sections. Triggered by the sparkle icon on any section in the Technical File tab. 20 calls per product per hour.
Obligation evidence (prompt key: suggest) – generates evidence narratives for each of 19 CRA obligations (Art. 13 through Art. 32). Same rate pool as technical file suggestions.
Vulnerability triage (prompt key: vulnerability_triage) – assesses CVE findings and recommends dismiss, acknowledge, or escalate with a confidence score. Available as single-finding or bulk triage (up to 20 findings per call). 5 calls per product per hour.
Incident report drafting (prompt key: incident_report_draft) – drafts ENISA notification content for the three CRA Article 14 reporting stages (24-hour early warning, 72-hour notification, 14-day final report). 5 calls per report per day.
Category recommendation (prompt key: category_recommendation) – analyses your product against CRA Art. 6–8 and Annex III to suggest the appropriate risk category (Default, Important I, Important II, or Critical). Uses Claude Opus for deeper reasoning. 5 calls per product per day.

Each capability uses only data already in CRANIS2: your SBOM, vulnerability scan results, repository metadata, stakeholder records, and existing documentation. The Copilot cannot access external sources or browse the internet.

Using AI suggestions

Select a station to begin