generated: '2026-08-13' method: probed source: https://stoplight.io/mocks/lucdya/Social-Listening-api/255000511 name: Lucidya Sandbox & Mock Surface type: Sandbox summary: >- Lucidya operates no test-mode environment, no test keys and no sandbox tenant. What does exist — and is genuinely callable without credentials — is the Prism mock server Stoplight stands up for each of Lucidya's five public API projects. Those mocks are reachable today and return the example payloads and example error bodies carried in Lucidya's own specs, which makes a gated API partially testable before a key is issued. vendor_sandbox: present: false test_mode: false test_keys: false test_key_prefix: null test_data_fixtures: false time_simulation: false note: >- Nothing in the developer documentation describes a sandbox, staging host, test-vs-live key distinction, or seeded test data. The only credential path published is a live production token issued after CSM approval. No test values are invented here. mock_servers: provider: Stoplight (Prism), auto-generated from Lucidya's own public Stoplight projects first_party: false first_party_note: >- These URLs are on stoplight.io, not on a Lucidya domain, and Lucidya does not link to them from its documentation. They are derived from the workspace Lucidya publishes (workspace slug "lucdya", domain docs.lucidya.com), so the example payloads they return are Lucidya's, but the hosting and the mock behaviour are Stoplight's. Recorded as a usable surface, not as a vendor-operated sandbox. authentication: none servers: - api: Social Listening url: https://stoplight.io/mocks/lucdya/Social-Listening-api/255000511 probed: '2026-08-13' example_request: GET /monitors_list?page_id=1 http_status: 400 response: '{"error":{"status":400,"detail":"Page_id is required"}}' note: >- Live and answering with the spec's own example error body. Prism returns the first declared response by default; a specific status can be selected with the `Prefer: code=200` header where the spec declares an example. - api: AI url: https://stoplight.io/mocks/lucdya/ai-api/254922095 probed: '2026-08-13' example_request: POST /sentiment/predict_batch http_status: 422 response: '{"detail":[{"loc":["string"],"msg":"string","type":"string"}]}' - api: CDP url: https://stoplight.io/mocks/lucdya/cdp-api/385761530 probed: '2026-08-13' example_request: GET /profiles http_status: 400 - api: OmniChannel url: https://stoplight.io/mocks/lucdya/omnichannel-api/400049121 probed: '2026-08-13' http_status: null note: Mock URL published by the project node; not exercised on this pass. - api: OmniServe Analytics url: https://stoplight.io/mocks/lucdya/omniserve/290278 probed: '2026-08-13' example_request: GET /analytics/pages http_status: 401 response: '{"status":401,"message":"Authentication required"}' note: >- The only mock that enforces the securityScheme, because OmniServe is the one spec that actually declares `OmniserveToken` on its operations. The other four leave securitySchemes empty, so their mocks answer anonymously. what_this_is_good_for: - Verifying error-envelope handling against Lucidya's real declared shapes before holding a key. - Confirming request/response schemas while a CSM access request is pending. - Contract-testing an integration without consuming production rate limit. what_this_is_not: - Not real data — Prism replays spec examples, it does not query Lucidya. - Not vendor-supported; Stoplight may retire a mock when a project is republished. - Not a substitute for a test tenant; there is no way to exercise writes, webhooks or job polling end to end. recommendation: >- The lowest-cost sandbox improvement available to Lucidya is to declare the `luc-authorization` securityScheme on all five specs (only OmniServe does today) and to link the mock URLs from the Get Started pages. Both are free — the infrastructure already exists and already runs. cross_links: authentication: ../authentication/lucidya-ltd-authentication.yml errors: ../errors/lucidya-ltd-problem-types.yml