Update to Luasai

{
  "description": "Populate or reconcile a spreadsheet, CSV, CRM export, or pasted account list with governed Luasai organization profiles and partner relationships. Use for multi-row account enrichment and writeback; not for one-company graph questions, internal data ingestion, or canonical data repair.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 578
    },
    {
      "relative_path": "references/account-list-workflow.md",
      "size_in_bytes": 56371
    },
    {
      "relative_path": "references/bounded-account-list-workflow.md",
      "size_in_bytes": 9997
    },
    {
      "relative_path": "references/identity-and-source-boundaries.md",
      "size_in_bytes": 46910
    },
    {
      "relative_path": "references/relationship-writeback-stage.md",
      "size_in_bytes": 9354
    }
  ],
  "name": "luasai-account-list-enrichment",
  "skill_md_contents": "---\nname: luasai-account-list-enrichment\ndescription: Populate or reconcile a spreadsheet, CSV, CRM export, or pasted account list with governed Luasai organization profiles and partner relationships. Use for multi-row account enrichment and writeback; not for one-company graph questions, internal data ingestion, or canonical data repair.\nmetadata:\n  short-description: Enrich account lists with Luasai intelligence\n  version: \"0.3.19\"\n---\n\n## Shared account recommendation precedence\n\nWhen `luasai_graph_recommend_accounts` is advertised, use it for the identity stage instead of client-owned candidate selection in the references below. Send every physical row's stable `input_id`, exact supplied `name`, its own `known_domains`/`aliases` and explicit `hq_country` in batches of at most 20, at most two in flight. Market is not HQ. Preserve the authoritative matched `organization_id`; a completed `no_match` is bounded, while `execution_incomplete` must be retried. Do not silently replace a recommendation with a candidate or parent. Retain the decision id and reason. Then use the existing relationship/reveal/writeback stage. Existing candidate-only workflows remain available when this capability is absent or an operator explicitly investigates. This paragraph takes precedence over their client-selection instructions.\n\n# Luasai Account Intelligence\n\nUse the Luasai MCP, and — only when an artifact is actually processed or\nwritten — the customer's approved file or spreadsheet capability. The Luasai\ntools supply governed organization and relationship intelligence; the client\nagent selects identities, applies the user's partner criteria, and writes the\nclient-owned artifact when a written output is requested.\n\nThis Skill is instructions and references only. It does not contain a Luasai\nclient, credentials, executable, hook, or local MCP server. For multi-row\ndiscovery every successfully completed row ends in exactly one state: **Matched\n— [platform account] [scope]**, or **No match — [reason]**. There is no terminal\nreview state. A transport, authorization, upstream, or client-state failure\ntakes precedence and stays incomplete; it is excluded from the completed totals,\nand the run is not complete until it is recovered.\n\n## Stages\n\nLoad only the stage you are executing, so a single-identity lookup stays small.\n\n1. **Identity stage (default).** Every account lookup begins here. Read and\n   follow\n   [references/bounded-account-list-workflow.md](references/bounded-account-list-workflow.md)\n   before you make any Luasai candidate or identity-lookup tool call — not\n   merely before choosing an organization. It is the complete default identity\n   reference: the physical-cell and stable-`input_id` rules, the typed-clue and\n   `alternate_query` rules, the one bounded preparation call and its fixed\n   limits, the one-batch own-ID-or-abstain selection, the single targeted detail\n   round, the source/tenant/identity boundaries, the binary outcomes, and the\n   execution-incomplete split all live there. Freeze one platform account ID or\n   one no-match per physical row before any later stage. If this module cannot\n   be loaded, the identity work stays incomplete and is never a No match.\n2. **Relationship, reveal, and writeback stage (only when requested).** When the\n   user has actually asked for partner relationships, a paid reveal, or a\n   written output, additionally read and follow\n   [references/relationship-writeback-stage.md](references/relationship-writeback-stage.md).\n   An identity-only lookup does not read it.\n\nThe longer identity files (`account-list-workflow.md` and\n`identity-and-source-boundaries.md`) are operator-only. They document the\nexplicit lower-level `luasai_graph_organization_candidates` investigation and\nits bulk/semantic recovery; they are not part of the normal default load.\n\n## Required capabilities\n\nLoad only the capabilities the operation you are performing actually needs.\n\n- **Identity preparation (identity stage, default).**\n  `luasai_graph_prepare_organization_choices` is the supported normal\n  preparation operation and is required. Preflight it once per run; when it is\n  missing, the identity work stays incomplete and never silently degrades into\n  an open-ended candidate research loop.\n  `luasai_graph_organization_candidates` is retained as the lower-level tool for\n  an explicit operator investigation, not the normal research loop.\n  `luasai_graph_organization_identity_evidence` is used only when the active\n  client advertises it and a candidate's identity is genuinely unresolved or a\n  needed fact is still omitted, as described below.\n- **Opening, processing, or writing a file or spreadsheet.** The provider's\n  approved native artifact workflow is required whenever a file or spreadsheet\n  is opened, processed, or written — including identity-only sheet work. It is\n  not required for a chat-only lookup.\n- **Chat-only lookup.** A lookup answered from a pasted list with no file\n  opened needs no artifact workflow.\n- **Requested relationship retrieval.** `luasai_graph_search` is required only\n  when the user requests partner relationships.\n- **Saved relationships and reveal.** `luasai_graph_revealed_relationships` is\n  used only when saved results may exist, and `luasai_graph_reveal` only for the\n  relationships actually needed, after the server quote and the existing client\n  approval flow.\n- **Identity-results-only writeback.** Writing identity results needs the\n  approved artifact workflow but not `luasai_graph_search`,\n  `luasai_graph_revealed_relationships`, or `luasai_graph_reveal`.\n\n`luasai_graph_report_data_issue` is an optional QA follow-up. If a capability an\noperation actually needs is unavailable, complete only the supported work and\nreport the exact blocker. Do not replace Luasai with a database query,\nrepository data, or web research.\n\n## Normal workflow (bounded preparation)\n\nThe supported normal path is short and lives in\n`references/bounded-account-list-workflow.md`; read it in full before the first\ntool call. In outline: read the source artifact once, preserving every tab and\nphysical ID; extract only the literal supplied clues into the row's typed\nfields; preserve the original names; send a bounded source-grounded\n`alternate_query` with basis `name_variant` only when a compound or overlong\ncell needs it; submit at most 20 rows per call to\n`luasai_graph_prepare_organization_choices`; select exactly one returned\n`organization_id` or an abstention per physical row in a single batch from your\nown facts and the source context, never treating the server's retrieval score\nas confidence; spend at most one targeted detail round when a decision genuinely\nneeds one omitted identity fact; and keep an execution-incomplete row separate —\nit is never a No match by default. The preparation returns a bounded shortlist\n(at most three candidates per executed stage), never a full-corpus search or an\nabsence proof.\n\n## Operator legacy path\n\nThe lower-level `luasai_graph_organization_candidates` whole-sheet bulk, per-row\nfast, and batched semantic investigation — with its mode caps, page limits, and\nper-row recovery budget — is documented in the operator-only\n`account-list-workflow.md`. Use it only when an operator deliberately needs that\nhands-on inspection. The normal path above submits each physical row once to\nthe preparation operation, which runs the same bounded retrieval and recovery\ninternally.\n\n## Optional identity-detail read\n\nWhen the active client advertises `luasai_graph_organization_identity_evidence`\nand one candidate's identity is genuinely unresolved, or a needed identity fact\nis still omitted from the returned profile, you may read that one candidate's\nomitted public identity evidence by passing back only the\n`identity_evidence_token` returned on that candidate row. The tool accepts no\narbitrary organization id, query, offset, or page: the token pins one\norganization and the requesting principal, tenant, and authentication type,\nexpires within ten minutes, and cannot be refreshed. Use it only for that one\ndecision, never to pre-fetch or page evidence for rows you can already decide.\nOne physical row's legacy rich-detail requests and its identity-detail reads\nshare the same two-read allowance, and every read shares the same at-most-two\ncalls in flight. A missing, rejected, changed, or failed necessary read is an\nincomplete execution, never a no-match. This read is internal resolver/agent\nworking data: the customer-facing result stays the matched platform account and\nits meaningful scope, or a no-match reason, and sources, citations, content\ndigests, capability tokens, and diagnostic receipts are never added to customer\noutput or shown as an evidence packet, source report, or audit tab.\n\n## Non-negotiable boundaries\n\nThe identity guide states the identity rules in full. The entry-level points\nthis router must itself carry are:\n\n- Preflight only the capabilities the operation needs. The preparation tool is\n  required for the normal identity lookup; the candidate tool is required only\n  for an explicitly requested operator investigation. Whenever a file or\n  spreadsheet is opened, processed, or written — including an identity-only\n  sheet pass — preflight and use the provider's approved native artifact\n  workflow, and stop before retrieving organization data if a needed capability\n  is missing. A chat-only lookup answered from a pasted list needs no artifact\n  workflow. Load and use the provider's approved spreadsheet or file workflow\n  before authoring; when a native artifact runtime is mandated, do not\n  substitute a generic library.\n- Keep the complete row and conversation context in the client agent, and send\n  only the typed hints published by the candidate tool. Assign a stable\n  physical-row ID to every requested cell before calling Luasai; a supplied\n  parent, group, or other column value is its own ordinary lookup with its own\n  `input_id`, never a special parent mode and never an ownership fact.\n- Never call or recreate the internal `/publicv2/ai/orgs/resolve/batch` lane;\n  candidate search is the interactive identity-discovery path. Never treat\n  organization roles as exclusive.\n- In Codex, invoke the native `spreadsheets:Spreadsheets` workflow for `.xlsx`\n  tasks and use `@oai/artifact-tool` for authoring, export, rendering, and\n  reread. `openpyxl`, LibreOffice, or another generic workbook library does not\n  satisfy Codex completion. If the Skill or runtime is missing, stop at the\n  preflight, and require an actual rendered-image review before completion.\n- For requested artifact delivery, never claim completion from a tool trace\n  or draft workbook alone. Delivery requires saved-artifact reread, row\n  reconciliation, rendered-image review, and repair followed by a second\n  render and review when any visual defect is found. A chat-only identity\n  lookup completes with reconciled physical-row decisions and private client\n  decision receipts; it does not create an artifact.\n"
}