Update to OloLand DD

{
  "description": "Use when the user asks to run the OloLand ololand-dd-managed-agent workflow. Run one of OloLand's cloud-hosted deal specialists on a deal — risk mapping, forensic QoE, pre-LOI screening, financing prep, IC memo drafting, or the multi-agent IC coordinator that runs risk + forensic + memo specialists in parallel. Each specialist reads the deal through OloLand's scoped, audited tool rail and answers with citations; the run is dual-written to OloLand's audit trail (agent_runs + OloLand's hosted agent service session id) so the result is provenance-complete. Asynchronous: starts the run, then polls to completion.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 451
    }
  ],
  "name": "ololand-dd-managed-agent",
  "skill_md_contents": "---\nname: ololand-dd-managed-agent\ndescription: \"Use when the user asks to run the OloLand ololand-dd-managed-agent workflow.\\\n  \\ Run one of OloLand's cloud-hosted deal specialists on a deal \\u2014 risk mapping,\\\n  \\ forensic QoE, pre-LOI screening, financing prep, IC memo drafting, or the multi-agent\\\n  \\ IC coordinator that runs risk + forensic + memo specialists in parallel. Each\\\n  \\ specialist reads the deal through OloLand's scoped, audited tool rail and answers\\\n  \\ with citations; the run is dual-written to OloLand's audit trail (agent_runs +\\\n  \\ OloLand's hosted agent service session id) so the result is provenance-complete.\\\n  \\ Asynchronous: starts the run, then polls to completion.\"\n---\n\n\n# Workflow for ololand-dd-managed-agent\n\nFollow the OloLand command instructions below when the user asks for `ololand-dd-managed-agent` or the equivalent workflow.\n\n# ololand-dd-managed-agent (cloud deal specialists)\n\nRuns one of OloLand's **managed deal specialists** — agents hosted on the OloLand's hosted agent service that work a deal through OloLand's scoped, audited tool rail and\nreturn a sourced, citation-bearing answer. Use it when a question deserves a\ndedicated specialist pass rather than an inline answer: a full risk read, a\nforensic QoE battery, a pre-LOI screen, financing prep, or an IC-grade\nsynthesis where the coordinator delegates to risk, forensic, and memo\nspecialists in parallel.\n\n## Usage\n\n```\nololand-dd-managed-agent <deal_id> <specialist> <question or instruction>\nololand-dd-managed-agent list\n```\n\n## Arguments\n\n- `deal_id` (required) — the OloLand deal the specialist should work on.\n- `specialist` (required) — one of the `template_id`s returned by\n  `list_managed_agents`. Typical roster:\n  - `risk-chat` — maps findings to the risk taxonomy with severity, likelihood, evidence\n  - `forensic-chat` — Beneish, Benford, EBITDA bridge, cross-document reconciliation\n  - `screen-chat` — pre-LOI go/no-go: premium analysis, accretion/dilution, antitrust HHI\n  - `financing-chat` — financing analysis, capital-provider sourcing prep, lender-package prep\n  - `ic-memo` — IC memo drafting; can persist a DRAFT IC package\n  - `ic-coordinator` — multi-agent IC synthesis (delegates to risk + forensic + memo in parallel)\n  - `lead-orchestrator` — general sourced deal Q&A\n- `question` (required) — what the specialist should do.\n\n## Execution\n\nThis is an asynchronous managed-agent run. Do not expect an immediate answer;\nstart it, then poll.\n\n1. **`list` mode.** If the user passed `list` (or asked what specialists\n   exist), call `list_managed_agents` and render each agent's\n   `template_id`, description, and whether it is `configured` in this\n   environment. Recommend the best fit for what the user is working on, then\n   stop.\n2. **Start.** Call `run_managed_agent` with `deal_id`,\n   `template_id`, and `prompt`. It returns a `task_id` (with\n   `task_type: \"managed_interactive_agent\"`).\n   - `{\"error_code\": \"not_found\"}` — wrong deal id; confirm it with\n     `list_deals`.\n   - `{\"error_code\": \"invalid_template\"}` — the specialist name is not\n     callable; show the roster from `list_managed_agents` and let\n     the user pick.\n   - `{\"error_code\": \"not_configured\"}` — the specialist exists but is not\n     provisioned in this environment; relay that and suggest another\n     specialist from the `list` output that shows `configured: true`.\n   - `{\"error\": \"Company scope required …\"}` — the MCP connection is not\n     company-scoped; tell the user to reconnect / sign in, do not retry\n     blindly.\n   - Any other tool-level error fails closed — relay it rather than\n     synthesizing an answer.\n3. **Poll.** Call `check_task_status` with the `task_id` every\n   few seconds. While `state` is `STARTED`/`PROGRESS`, relay the progress\n   message. The IC coordinator's delegating turns run multi-minute — keep the\n   user informed rather than giving up. Keep polling until `state` is\n   `SUCCESS` or `FAILURE`.\n4. **On `FAILURE`** — report the `error` verbatim. Do not fabricate a\n   specialist answer.\n5. **On `SUCCESS`** — the `result` object is:\n   - `response` — the specialist's answer, already source-cited. This is the\n     deliverable; render it unaltered.\n   - `run_id` — the OloLand `agent_runs` audit row.\n   - `claude_platform_session_id` — the hosted session for operator trace.\n   - `tool_calls`, `input_tokens`, `output_tokens` — run telemetry.\n   - `citations` — structured citation payloads backing `[N]` markers, when\n     present.\n\n## Output\n\nPresent the specialist's `response` unaltered, then append:\n\n```\nSpecialist run — deal: <deal_id>  specialist: <template_id>\nProvenance: run_id=<run_id>  session=<claude_platform_session_id>  tool_calls=<tool_calls>\n```\n\n## Honest bounds\n\n- The specialist reads the deal only through OloLand's scoped tool rail — it\n  answers from the ingested corpus and deterministic engines, and says so when\n  evidence is missing. It cannot approve anything: IC package approval, patch\n  approval, and version commits are human-only actions in the OloLand app.\n- `ic-memo` / `ic-coordinator` may persist an IC package **draft**; approval\n  stays in-app.\n- `financing-chat` prepares analysis and sourcing materials only; it never\n  spends Apollo credits, sends email, or submits lender applications.\n- If a run reports no usable evidence, relay that — a specialist run is not a\n  substitute for ingesting the data room first (`ololand-dd-new-deal`, document upload).\n\n## Example\n\n```\nololand-dd-managed-agent deal_acme_2026 forensic-chat Run the full QoE battery and flag anything that moves adjusted EBITDA by >5%.\n```\n\nCompanion: `/conflicts` (unattended cross-document conflict scan), `ololand-dd-verify`\n(per-figure corpus verification), `ololand-dd-dd-analyze` (bounded extraction, risk, and\nfinancial-snapshot pipeline; reconciliation is a separate explicit action).\n\n\n## OpenAI execution requirements\n\nUse the connected MCP server's current tool schemas. Do not invent tool arguments, job-status endpoints, inputs, or outputs. If a referenced tool is unavailable, explain the limitation and request the needed connection or data. Invoke writes only when the user has requested the corresponding action. Do not initiate purchases or payments. An unrun, unsupported, or unavailable procedure is a gap, never a passing result.\n\n\nOpenAI cannot execute local lifecycle hooks. Use the packaged explicit setup/review workflow when requested. Server-side MCP auditing remains authoritative; preserve evidence blockers and human-only approval gates.\n\nClaim-correction submission is unsupported in this OpenAI upload because the current handler records surface=cowork. Do not invoke submit_agent_claim_correction from OpenAI. Explain the limitation and retain proposed correction text for human review without claiming submission. See unsupported-workflows.json for the excluded canonical workflow.\n"
}