Update to Vertical Bar Agent

{
  "name": "stress-test",
  "description": "Author, revise, preview, publish, run, inspect, and stop CrossCheck Stress Tests against NetSuite. Use for load curves, traffic-volume experiments, Stress Test contracts, smoke/full runs, run progress, results, or cancellation. Agent execution is limited to verified sandbox or release_preview environments; hand production and development execution to the user in CrossCheck. Accepted runs are asynchronous with an immediate Run link. For regression assertions without load use test-suite; for writing application unit tests use neither.",
  "included_files": [
    {
      "relative_path": "references/authoring.md",
      "size_in_bytes": 3283
    },
    {
      "relative_path": "references/operations.md",
      "size_in_bytes": 4133
    }
  ],
  "skill_md_contents": "---\nname: stress-test\ndescription: Author, revise, preview, publish, run, inspect, and stop CrossCheck Stress Tests against NetSuite. Use for load curves, traffic-volume experiments, Stress Test contracts, smoke/full runs, run progress, results, or cancellation. Agent execution is limited to verified sandbox or release_preview environments; hand production and development execution to the user in CrossCheck. Accepted runs are asynchronous with an immediate Run link. For regression assertions without load use test-suite; for writing application unit tests use neither.\n---\n\n# Stress Test — from workload intent to a live Run\n\nThe platform owns execution and evidence. You turn the user's workload into a reviewed contract,\nthen perform only the operations they requested. Authoring does not imply execution. Smoke writes\nreal NetSuite records too; neither mode cleans them up.\n\n## Route the intent first\n\n| Request | Route |\n| --- | --- |\n| Create or revise a contract | [Authoring](references/authoring.md): discover, resolve cases, draft, validate and preview, review, persist |\n| Publish | Read the exact revision, review its digest and effect, publish |\n| Run | [Operations](references/operations.md): verify sandbox/release_preview; hand production and development to the user in CrossCheck, otherwise review and start asynchronously |\n| Progress or results | List runs if needed, then read status or bounded report; no mutation approval |\n| Stop | Resolve one exact Run, request cancellation; explain acknowledgement versus completed stop |\n\nResolve `workspaceId` with `cc_workspaces` and the target environment with `cc_environments`.\nReuse an unambiguous selection already made in this conversation. Ask only when multiple candidates\nremain; never choose the first match or invent identifiers. Use the same authorized workspace on\nevery related call. Do not copy a Run reference from a different workspace.\n\nBefore each logical mutation, summarize target, concrete change and effect once. If the user already\nauthorized that exact scope, proceed; otherwise ask once. Creating a contract never authorizes a run.\nReady + publish is one operation. An explicit stop request authorizes stopping the identified Run;\ndo not add another approval loop. Core authorization remains authoritative.\n\n## Production and development execution belong to the user in CrossCheck\n\nBefore any start, retry or rerun, call `cc_environments` for the selected workspace and match the\nexact environment ID. Only an explicitly returned `environmentType` of `sandbox`\nor `release_preview`, with `isActive: true`, permits Agent execution. Names, account-ID patterns,\nuser assurances and the MCP server's own\ndeployment environment are not proof. Missing, conflicting or unknown classification means do not\nstart; resolve it first. Do not relabel an environment to make execution eligible.\n\nFor `production` or `development`, refuse both smoke and full starts even when the user approves or requests an\nexception. Explain that they must open the Stress Test in CrossCheck and start the Run themselves\nusing the product's confirmation flow. Do not call another execution tool, delegate the start,\nprovide an executable API workaround, or click the product's Run controls on their behalf.\nDo not substitute a different environment without the user's selection.\n\nAuthoring, validation, preview, publishing, status/report reads and an explicitly requested stop\nremain available for production and development. No Run is created by the handoff, so do not invent a Run URL.\nAfter the user starts it, resolve the actual Run and return its server-provided link. The MCP runtime\nindependently verifies fresh authorized environment metadata before every start\nand rejects production, development or unverified targets before dispatch. This also applies to Test Suite and\nCI workflow starts. Product UI execution remains available; the skill must not click Run for the user.\nThis gate takes precedence over retry guidance: if a start timed out and the refreshed type is\nproduction, development or unknown, inspect existing Runs to reconcile acceptance without resending the start.\nKeep the original idempotency key; an uncertain response never proves that no Run was created.\n\n## Non-negotiable behavior\n\n- Arrival rate is the default authoring model. Ask for traffic volume and duration; don't ask for\n  concurrency unless the user explicitly needs a concurrent-worker experiment.\n- Read `cc_stress_test_capabilities` before authoring. Flows reference published canonical Test Suite\n  cases; no scripts, duplicated step language, guessed record IDs, or guessed unsupported actions.\n- Preserve unrelated flows, pools, stages and tags when editing. Use the current definition digest\n  for updates and the exact reviewed digest for publishing.\n- Show whether variables come from supplied rows, generated values, or verified account reads.\n  Generated business IDs are not verified IDs. Do not put secrets or credentials in data pools.\n- Generate a unique idempotency key for each new mutation intent and retain it. Reuse it after a\n  timeout or uncertain response; never turn a retry into a new run. Changed intent needs a new key.\n- After start acceptance, immediately show the server's **runUrl** as a clickable link and identify\n  environment, revision and smoke/full mode. Say \"accepted\" or the returned state, not \"finished\".\n  End the response without waiting for completion. Poll only if the user asked you to watch; respect\n  `suggestedPollAfterSeconds`, and still send the link before polling.\n- Use returned **runUrl/resultUrl** verbatim. Never compose URLs from guessed slugs, internal IDs or\n  the MCP host. If the server supplies no link, say it is unavailable and preserve the Run reference.\n- A terminal Run may still be collecting its final report. Report these states separately. Missing\n  account observations are unavailable/unobserved, not zero. There is no performance pass/fail verdict.\n- Do not introduce schedules, automatic stops, thresholds, automatic cleanup, or extra backend\n  approval fields. Do not silently lower, raise, or reinterpret a requested load curve.\n- On a refusal, report the code and actionable paths/reasons. Read and repair a conflicting draft;\n  do not overwrite or publish a newer definition under an older authorization.\n"
}