sentinelone v1.0.0

The pack names a reference, never a secret. Installing a pack registers capabilities and makes them governable; registration is not connectivity. The credential is created at the provider and held in the client's own secret manager under the reference the pack names, and deployment testing is still required to show the credential path works and cannot be bypassed.

Reference

secret-ref-sentinelone-api-token

Custody

customer_vault

Injection

header (Authorization)

A SentinelOne console-issued API token, held in the client's own secret manager. The stored material must be the complete Authorization header value: the scheme 'ApiToken', one separating space, then the token — 'ApiToken <token>'. Token issuance is an external process: an admin generates the token in the management console for a console user (use a dedicated service user), with whatever expiry options the tenant's console version offers (provisional — confirm at generation), and stores the value under this reference; the executor injects the stored value verbatim and performs no token lifecycle itself, so rotation ownership sits with that external process. A console API token inherits the full permissions of the generating user's console role and scope (account/site) — create it from a dedicated least-privilege service user: a Viewer-scoped user for the read capabilities, and an IR/Admin-scoped user only where the response actions (isolate, mitigate, resolve) are enabled, ideally as two tokens under two references. Expiry and revocation statements here are PROVISIONAL, not verified: the console-gated API Hub documentation could not be retrieved (docs.sentinelone.com was unreachable again on 2026-09-21), so the deployment must confirm against the tenant's own console version before enabling — whether token expiry is optional or mandatory for that version (earlier drafts of this note asserted both an optional expiry and a universal validity period, an unreconciled contradiction), how expiry is displayed at generation, and whether regenerating or revoking a token invalidates the old value immediately. Until confirmed, treat any copy of the token captured before revocation as live: revocation certainty depends on the console version, and credential rotation on suspicion must not assume instant invalidation; an expired or rejected token fails closed. Executor hygiene: the value must live only in executor memory for the lifetime of a call, must never be written to logs, traces or diagnostics, and must not appear in crash dumps; a deployment that cannot meet that bar must not enable the pack.

• sentinelone-agents (api, pii): https://sentinelone-console.invalid/web/api/v2.1/agents
• sentinelone-threats (api, internal): https://sentinelone-console.invalid/web/api/v2.1/threats
• sentinelone-console (api, internal): https://sentinelone-console.invalid/web/api/v2.1

Source disclosure: SentinelOne's official API reference is console-gated — the vendor's own FAQ page (the pack's one retrieved source) states that API documentation is available directly in the SentinelOne Ops Center under Help → API Hub as OpenAPI/Swagger, so no public operation page could be retrieved for any registered endpoint. The registered paths follow the long-standing, widely-corroborated /web/api/v2.1 route shapes, and every capability must be re-verified against the tenant console's API Hub before this pack leaves draft. The endpoint host 'sentinelone-console.invalid' is a reserved non-production placeholder (an RFC 2606 reserved TLD): substitute the customer's SentinelOne management console host (the same console base URL that serves their /web/api/v2.1 API) before enabling, in every endpoint, in the resource locators and in every egress allowlist entry — the allowlist matches exactly, so all three must be substituted together, and substitution is a required deployment step. The pack does not create the console service user or the API token: an external deployment process generates the token in the management console and stores the complete header value described in the credential note. Endpoint isolation and threat mitigation are registered as the privileged response actions they are: agents.isolate disconnects an endpoint from the network (deploy at the R3 floor as a response-action equivalent), threats.mitigate executes kill/quarantine/remediate/rollback on the endpoint (R3), and threats.resolve suppresses a threat from the active queue (a reversible status change at the R2 update floor, with the suppression risk stated in its description and bounded by policy). Documentation caveat: SentinelOne serves its interactive API documentation inside the management console behind a customer login — the documentation URLs attempted in this review returned only the console's login shell, and docs.sentinelone.com was unreachable when tried again on 2026-09-21, so no vendor operation page could be read for this pack and no tenant-version API Hub evidence could be added to the review sources; the FAQ page in the review sources is the only retrieved page and supports ONLY the documentation-gating claim — it establishes no endpoint, request schema or authentication behavior. The endpoint paths and request shapes here reflect the widely used public v2.1 API surface as documented by SentinelOne in-console and in community SDKs, but a completing review MUST re-verify every path, verb and body shape against the in-console API documentation of the customer's own console version before this pack leaves draft, and must confirm the response-action semantics (mitigate actions, disconnect/connect) against that version. Documented exclusions — this pack does not register: agent uninstall and decommission (irreversible removal of protection from the endpoint), remote shell / RemoteOps script and command execution (arbitrary code execution on endpoints), threat rollback as a standalone capability (system-state rollback is destructive) — note that excluding the standalone capability does NOT exclude the rollback effect: rollback IS included through the registered threats.mitigate ACTION selector, and only an actual ACTION binding or policy restriction on that registered capability removes it, initiate-scan and full-disk scans (endpoint-disruptive), policy, exclusion and blocklist/hash management, site and account structure management, console user and role management, and ranger/network-quarantine configuration.

Draft review record dated 2026-09-21 · Reviewer: aaes-eng (an opaque id that resolves internally to the responsible reviewer)

• https://www.sentinelone.com/faq/ (retrieved 2026-09-21)

A pack review is not a vendor or examiner attestation. A retrieval date says when a source page was read; it does not by itself establish what the page said then. No pack is vendor-endorsed, and none has been connectivity-tested against a live vendor tenant.