Release walkthrough archive

LIQUILENS · PLUGIN REVIEW EVIDENCE

Seiche

Runtime 0.16.10 · Plugin 1.0.3 · Captured 29 October 2026. Five successful public MCP calls from the accepted production release.

90.0 seconds · edited browser recording · no audio · no model API calls for this recording.

Inspect the interactive capture viewer · Case transcript · Recording manifest · Underlying service captures

What this demonstrates

The video shows captured responses from the public MCP service, including the eight-tool release boundary, source clocks, withheld fields and coverage gaps. It is not a native plugin-installation recording or a new model-routing test.

Direct public MCP calls only. Product-skill routing and negative model-routing cases are review expectations, not newly executed model tests. No trading, enrollment, collection, private-record access or execution was performed. Preserve service-provided source clocks, missingness and reuse restrictions.

Cases shown

  1. Source rights and release scope

    Use Seiche to check which source-backed research tools are available and which legacy features are unavailable.

    Interpretation boundary: Report the eight public research tools and the returned source-rights restrictions. Explicitly identify legacy forecasts and backtests as unavailable; do not reconstruct withheld values from other sources or treat availability as investment approval.

  2. Money-market overview

    Use Seiche to summarize the money-market evidence available for a treasury review.

    Interpretation boundary: Request the summary section and explain source coverage, units and limits without treating transport success as validated current data.

  3. India funding context

    Use Seiche to review India's money-market context in a concise summary with source clocks.

    Interpretation boundary: Request the india section and summary view. Preserve missing observations and regional scope; do not infer institution-specific exposures.

  4. Source health before use

    Check Seiche's data health and identify sources that need follow-up before I use a funding number.

    Interpretation boundary: Report the returned source states and clocks. Distinguish healthy transport from data freshness and explicitly flag denied or stale evidence.

  5. GIFT City research context

    Use Seiche to prepare a short GIFT City research note with source links and coverage limits.

    Interpretation boundary: Return the available GIFT City context with source dates and reported limits, retaining unavailable values and avoiding investment or regulatory approval claims.

Endpoint: https://agents.seiche.info/products/seiche/mcp

Runtime source: 98359e71

Video SHA-256: faf6b2dbbc5f8ad935ffac28a7f9493069c7d163890e34418f8acf46627b3d8d

Review evidence. This page is not an OpenAI approval or publication receipt.