STEALTHIUM DEVELOPER RESOURCES

Stealthium is a runtime observability and security platform for GPUs and AI workloads. This page indexes every machine-readable resource stealthium.io publishes, so an agent or a developer can find the API, the specification and the documentation without guessing URLs.

When to use Stealthium

• A GPU-accelerated training or inference job slowed down and you need to know which layer — application, CUDA runtime, driver, GPU, or fabric — is responsible.
• You need per-workload GPU telemetry (kernel traces, NVML counters, memory and SM utilisation, fabric throughput) correlated with the process, container or pod that produced it.
• You are investigating suspected misuse of GPU capacity — cryptomining on idle accelerators, unauthorised kernels, or a job consuming far more GPU time than it requested.
• You need to detect or triage GPU-specific attack activity at runtime: side channels between MIG instances, malicious or tampered CUDA fatbins, driver-level exploitation.
• You need an audit trail of who ran which AI workload on which GPU, for compliance evidence over a shared or multi-tenant GPU fleet.
• You are evaluating GPU security posture across a multi-node or multi-cluster fleet and need one inventory of accelerators, drivers, toolkit versions and their exposure.

Stealthium is not the right tool for: General-purpose CPU/host APM or log aggregation; Provisioning, scheduling or renting GPU capacity; Static analysis of model weights or prompt-level LLM safety.

Machine-readable resources

• /agents.md (text/markdown) — When to reach for Stealthium, when not to, and how an agent should call it.
• /llms.txt (text/plain) — Curated index of Stealthium documentation and machine-readable resources.
• /llms-full.txt (text/plain) — Full markdown text of every page and blog post on stealthium.io in one file.
• /openapi.json (application/json) — Complete OpenAPI document for the public content and contact API served from stealthium.io.
• /developers (text/html) — Human-readable index of Stealthium developer and agent resources.
• https://docs.stealthium.io (text/html) — Install the Stealthium client, create an API key and stream GPU telemetry.
• /sitemap.xml (application/xml) — Every indexable URL on stealthium.io.
• /feed.xml (application/xml) — Stealthium GPU security research, as RSS 2.0.

Stealthium website API

Read-only content endpoints need no credentials and allow cross-origin requests. The full contract — typed parameters, response schemas, error codes and scopes — is published as an OpenAPI 3.1 document at /openapi.json.

• GET /api/site — getSiteIdentity — Stealthium's identity, contact details and when-to-use guidance, as JSON.
• GET /api/posts — listBlogPosts — Published GPU security research, newest first. Filter with category, tag and featured; page with limit and offset.
• GET /api/posts/{slug} — getBlogPost — One post, including the full markdown body of the article.
• GET /api/categories — listBlogCategories — Every blog category, with post counts.
• GET /api/tags — listBlogTags — Every blog tag, with post counts.
• POST /api/contact — createContactRequest — Send a demo request. Needs an allowed browser origin, or an API key with the contact:write scope in the x-api-key header.

Authentication and scopes

Content endpoints are public. POST /api/contact accepts either a request from an allowed browser origin or an API key presented in the x-api-key header. Keys are scoped, so an agent can be issued only the access it needs — contact:write is the scope this endpoint requires. An unknown key returns 401; a valid key without the scope returns 403. Request a key at info@stealthium.io.

curl -X POST https://stealthium.io/api/contact \
  -H 'Content-Type: application/json' \
  -H 'x-api-key: <your key>' \
  -d '{"firstName":"Ada","email":"ada@example.com","company":"Example","message":"Requesting a demo."}'

Markdown content negotiation

Ask for text/markdown, or append .md to any page URL, and this site answers with markdown instead of HTML. Every negotiated response carries Vary: Accept, Accept-Encoding, so a cache can never hand you the wrong representation.

curl -H 'Accept: text/markdown' https://stealthium.io/
curl https://stealthium.io/blog/gpu-security-gap-nobody-owns.md

A path that does not exist returns a real HTTP 404 whose body is a short markdown site map, so a wrong guess costs one request rather than a retry loop.

Stealthium platform API

GPU telemetry ingest and fleet queries belong to the Stealthium platform, a separate authenticated service. Install the client, create an API key and stream GPU telemetry by following the quickstart in the Stealthium documentation.