generated: '2026-07-19'
method: generated
source: >-
  Generated from openapi/ (itself generated from Leena AI's published prose documentation)
  plus conventions/, errors/, scopes/ and rate-limits/ in this repo. Leena AI publishes no
  Agent Skills or AGENTS.md of its own — searched 2026-07-19.
provider_published_skills: false
summary: >-
  Three packaged Agent Skills covering the marquee flows of Leena AI's external API surface:
  running an agent procedure, exporting audit logs, and syncing the knowledge base. Every
  operationId referenced is verified against the specs in openapi/.
skills:
  - name: leena-ai-run-agent-procedure
    file: leena-ai-run-agent-procedure.md
    description: >-
      Trigger an Agent Operating Procedure on an AI Colleague and poll it to a terminal
      state, handling the paused (human-approval) state and the absence of idempotency.
    api: leena-ai:external-aop-api
    spec: openapi/leena-ai-aop-openapi.yml
    operations:
      - executeAop
      - getAopStatus
    risk: high
    risk_note: >-
      executeAop starts an autonomous agent run that can act across connected enterprise
      systems, and carries no idempotency key.
  - name: leena-ai-export-audit-logs
    file: leena-ai-export-audit-logs.md
    description: >-
      Incrementally export audit records via cursor pagination for SIEM or warehouse sync,
      within the 60 rpm limit and handling the PII the records carry.
    api: leena-ai:audit-logs-external-api
    spec: openapi/leena-ai-audit-logs-openapi.yml
    operations:
      - listAuditLogs
    risk: low
    risk_note: >-
      Read-only, but the response carries actor email, phone, employeeId, IP and user agent.
  - name: leena-ai-sync-knowledge-base
    file: leena-ai-sync-knowledge-base.md
    description: >-
      Migrate or continuously sync knowledge articles and attachments into the Leena
      Knowledge Base with permission scoping and a final indexing trigger.
    api: leena-ai:knowledge-management-rest-connector
    spec: openapi/leena-ai-knowledge-management-openapi.yml
    operations:
      - createKnowledgeAccessToken
      - uploadKnowledgeAttachment
      - syncKnowledgeArticle
      - completeKnowledgeSync
    risk: medium
    risk_note: >-
      Writes the corpus that grounds agent answers. Idempotent by reference_id upsert, so a
      stable source identifier is essential.
not_covered:
  - what: MCP tool flows (send_message, request_details, cancel)
    why: >-
      Consumed through Leena AI's hosted MCP server by any MCP-capable client, so a packaged
      skill would duplicate the protocol's own tool descriptions. See mcp/leena-ai-mcp.yml.
  - what: Analytics querying (queryAicAnalytics)
    why: >-
      Single-operation surface with one documented metricId; covered adequately by the spec
      and the conventions note about its HTTP 200 soft-failure semantics.