privacy/workflows/transcribe.json

{
  "schema_version": 1,
  "workflow": "transcribe",
  "display_name": "Transcribe",
  "governed_paths": [
    "skills/transcribe/SKILL.md",
    "scripts/launch_hosted_voice.py",
    "scripts/upload_hosted_audio.py",
    "scripts/import_hosted_voice_bundle.py",
    "scripts/import_hosted_voice_bundle_to_folder.py",
    "scripts/import_latest_hosted_voice_bundle.py",
    "scripts/finalize_hosted_transcript.py",
    "scripts/integrate_transcript_review.py",
    "scripts/self_relaunch.py",
    "scripts/managed_python_runtime.py",
    "scripts/_managed_python_runtime.py",
    "requirements.txt",
    "components.json",
    "scripts/_shared_python_runtime.py",
    "requirements-shared-core.txt",
    "requirements-shared-ocr.txt",
    "constraints-shared-macos-py312.txt",
    "scripts/_python_bootstrap.py",
    "scripts/build_voice_feedback_timeline.py",
    "scripts/bounded_process.py"
  ],
  "codex_context": {
    "policy": "real_professional_data_may_enter_codex_context",
    "classes": [
      {
        "id": "recording-and-case-context",
        "purpose": "Transcribe an advisor debrief, meeting, call, or existing recording in the correct project context",
        "content": "Audio or screen-recording content; compact case brief; client and project facts; recording title, date, participants, interviewer, notes, source transcription language (including Arabic), separate supported case output language, and source metadata"
      },
      {
        "id": "transcript-and-review",
        "purpose": "Attribute speakers, review evidence, and integrate the transcript",
        "content": "Raw and reviewed transcripts, speaker identities or labels, timing and media metadata, uncertainty notes, transcription corrections, review findings, exact transcript artifact hashes, quotation evidence receipts, stable claims and dependency edges, judgement proposals bound to their direct evidence, questions, issues, and shared evidence- and claim-register updates"
      },
      {
        "id": "retained-converter-attempts",
        "purpose": "Inspect frame extraction and document conversion results or failures",
        "content": "Local feedback-frame extraction retains images and per-frame converter stdout/stderr logs in attempt directories, including after failure or cancellation. These files may contain screen-recorded professional content, media metadata and local paths. Their contents enter model context when read; generating or retaining them does not itself send them to Hosted Voice. They remain until separately removed."
      }
    ]
  },
  "ordinary_codex_model_processing": {
    "scope": "content_supplied_to_the_codex_model",
    "account_arrangement": "user_selected_chatgpt_or_codex_account",
    "separate_clara_recipient_or_arrangement": false,
    "automatic_anonymisation": false,
    "local_filter_or_aggregate": "only_when_useful_for_professional_work",
    "plan_visibility": "not_inspected_or_enforced_by_clara"
  },
  "codex_account_boundary": {
    "selected_by": "firm_or_user",
    "clara_runtime_enforcement": "none",
    "review_timing": "before_professional_use_and_when_account_or_terms_change",
    "review_items": [
      "account_or_workspace_plan",
      "model_training_data_controls",
      "retention_and_deletion_controls"
    ],
    "per_case_record_required": false
  },
  "hosted_service_ids": [
    "hosted-voice"
  ],
  "boundaries_beyond_codex": [
    {
      "id": "hosted-voice-capture-or-upload",
      "kind": "hosted_service",
      "hosted_service_id": "hosted-voice",
      "destination": "Mparanza Hosted Voice and its configured realtime/transcription provider",
      "purpose": "Capture or upload a recording and return a transcript bundle",
      "content": "Compact case context when enabled, microphone or uploaded audio, source metadata, and source transcription language reach Hosted Voice. Arabic source audio returns an Arabic transcript; the separate supported case output language controls later local notes, analysis, and deliverables. Screen video, active-slide timeline, and visual capture metadata remain in the browser and local downloaded bundle; the service returns transcript and transcription metadata.",
      "optional": true,
      "requires_confirmation": false,
      "controls": [
        "The route is used only when the user asks for hosted capture or upload; importing an existing local bundle does not call the service.",
        "The separate hosted-service record states the source-backed access and retention position.",
        "Authentication material is not stored in transcript or advisory artifacts."
      ]
    },
    {
      "id": "direct-cli-python-dependency-setup",
      "kind": "public_research",
      "destination": "Python Package Index (PyPI) or the index selected by the user's Python configuration",
      "purpose": "Prepare the published shared Vera, Clara and Lucia core dependencies, and validate the selected workflow, in one user-scoped Python 3.12 environment per operating-system host.",
      "content": "Shared published package names and version constraints plus ordinary package-index request metadata. Client files, prompts, case data and generated work are not included in installer requests. Workflow arguments stay in the local child process.",
      "optional": false,
      "requires_confirmation": false,
      "controls": [
        "Only published shared requirements are installed, never requirements derived from client material or prompts.",
        "One fixed environment outside plugin source and client folders is reused across products and modules. Explicitly approved OCR is retained in the same environment.",
        "A reader lease prevents setup from modifying packages while managed workflows run. Setup validates dependencies before writing the readiness receipt. Failed updates leave execution unavailable until repair.",
        "Older plugin policies cannot downgrade an environment created by a newer shared policy revision."
      ]
    },
    {
      "id": "declared-python312-retrieval",
      "kind": "public_research",
      "destination": "Astral python-build-standalone CPython distributions on GitHub, or the Python download mirror explicitly configured in uv",
      "purpose": "Provision the declared CPython 3.12 workflow interpreter when absent, automatically bootstrapping uv if needed",
      "content": "The fixed CPython 3.12 version request, operating-system and architecture selection, and ordinary download request metadata. Setup does not read client files or add prompts, case material or generated reports to this request.",
      "optional": false,
      "requires_confirmation": false,
      "controls": [
        "Use an installed CPython 3.12 when available; otherwise use uv or automatically download the published uv 0.12.10 wheel, verify its pinned SHA-256, and provision private CPython 3.12 without changing system Python or shell profiles.",
        "Probe the selected interpreter before creating the dependency environment; never fall back to executing workflows with another Python minor version.",
        "A failed download or interpreter probe stops setup; existing environments and client files remain intact."
      ]
    }
  ],
  "security_controls": [],
  "review": {
    "reviewed_at": "2026-09-15",
    "reviewed_by": "privacy-surface-review",
    "basis": "external_boundary_review_of_workflow_source",
    "source_fingerprint": "c4b9bcdc540047e5ff1ecadc3945f5aae29cc516f24598eac4079923c1f1d44e"
  }
}