/api/vision/capabilities
platphorm-catalog remains a public, read-only capability catalog. This selection maps existing repository capabilities to an operator workflow; it does not execute Browser, PlatPhorm Content, or PlatPhorm Docs actions from the public site.
Repositories
207
Capabilities
49
Tools
3
Gaps
0
Browser
Captures UI route evidence, screenshots, journey checks, accessibility checks, and public read-only product verification.
Screenshots and browser artifacts must be public-safe and redacted before publication; protected journey triggers require PLATPHORM_API_KEY.
Publishes operator-approved reports, screenshots, exports, and evidence packets through the PlatPhorm Content surface instead of scattering artifacts across local repos.
The public catalog only declares the platform fit. Content writes, sharing changes, moves, imports, and exports are protected operator actions.
Produces polished briefs, remediation reports, privacy-defense packets, and PlatPhorm Docs-ready deliverables from evidence-backed catalog data.
Generated documents must distinguish source evidence from recommendations and must not include secrets, raw IPs, cookies, or private client data.
Builds an exposure graph, scores exploitability, and creates prioritized OPSEC reduction actions from real source evidence.
Start with read-only footprint import, graph/risk summaries, and operator-reviewed remediation notes. Takedowns and client data ingestion stay protected.
Turns public findings into risk-ranked tasks, evidence bundles, and tracked remediation workflows without pretending removal succeeded.
Expose discovery and task generation as public-safe previews. Real takedown, replay, publish, or workflow-run actions require PLATPHORM_API_KEY and legal review.
Converts exposure signals into operator-readable guidance, compliance status, and next actions tied to evidence.
Keep scoring explainable and evidence-backed. Avoid behavioral surveillance claims until client-side consent, storage, and retention controls exist.
Uses the PlatPhorm standard routes, discovery files, and best-implementation rankings to prove each site remains human-usable and machine-readable.
Public reads can show status and gaps. Regeneration, scans, content writes, docs publication, and remediation are protected actions.