mirror of
https://github.com/hanzoai/insights.git
synced 2026-11-03 00:52:32 +00:00
zeekay
1b367e2b62
The tree was a debranded snapshot of upstream taken around 2026-02-20, and upstream has moved about five months since. This replaces it with upstream master, mechanically re-derived: bin/debrand does the rename, and the pipeline and squashed migrations are restored from the shipped tree. Not a merge. The previous import replaced the tree wholesale in a single commit, so there is no shared content history to merge against. Reproducing the rename on the new tree is what actually converges. What survives: .hanzo/workflows, the Dockerfiles and bin/docker* entrypoints, compose files, LLM.md, and insights/migrations. The squashed initial migration is load-bearing — the live database records it by name, so upstream's ~918 individual migrations cannot come back. Nine products are dropped as unreferenced. Twenty-one more that we do not run are kept anyway because the core app still imports them; unpicking those is a refactor and does not belong in a rebase. This base does not build yet, and the reasons are not in the rename: upstream now wants Python 3.13.13 against our 3.12.12 and Django 5.2 against our 4.2, and the dependency manifests still name upstream packages that we republish under our own names. Those are decisions, not renames.
121 lines
4.5 KiB
YAML
121 lines
4.5 KiB
YAML
# OpenAI endpoints that are NOT on the HIPAA-eligible list per
|
|
# https://cdn.openai.com/osa/hipaa-endpoints.pdf:
|
|
# - Assistants API (/v1/assistants)
|
|
# - Threads API (/v1/threads)
|
|
# - Fine-tuning (/v1/fine_tuning/*)
|
|
# - Realtime API (/v1/realtime)
|
|
# - Legacy text completions (/v1/completions — distinct from /v1/chat/completions)
|
|
#
|
|
# Anything Insights routes through the OpenAI SDK must hit only HIPAA-eligible
|
|
# endpoints while we are under the BAA. New code that needs one of the above
|
|
# must be gated off the HIPAA-customer surface OR refactored onto an eligible
|
|
# endpoint (chat completions, responses, embeddings, audio, moderation, ...).
|
|
#
|
|
# If a non-PHI flow legitimately needs an ineligible endpoint, suppress per-line
|
|
# with `# nosemgrep: <rule-id>` and a short reason.
|
|
rules:
|
|
- id: openai-non-hipaa-assistants-threads
|
|
patterns:
|
|
- pattern-either:
|
|
- pattern: $X.beta.assistants
|
|
- pattern: $X.beta.threads
|
|
- pattern: $X.assistants.$M(...)
|
|
- pattern: $X.threads.$M(...)
|
|
message: |
|
|
OpenAI Assistants / Threads APIs (`/v1/assistants`, `/v1/threads`) are
|
|
NOT on the HIPAA-eligible endpoints list. Use the Responses API
|
|
(`client.responses.create(...)`) or Chat Completions
|
|
(`client.chat.completions.create(...)`) instead.
|
|
|
|
For non-PHI flows that nonetheless need Assistants/Threads, add
|
|
`# nosemgrep: openai-non-hipaa-assistants-threads` with a reason.
|
|
languages: [python]
|
|
severity: ERROR
|
|
metadata:
|
|
category: compliance
|
|
subcategory: hipaa
|
|
confidence: HIGH
|
|
paths:
|
|
exclude:
|
|
- '**/.semgrep/**'
|
|
- '**/test/**'
|
|
- '**/tests/**'
|
|
- '**/test_*.py'
|
|
|
|
- id: openai-non-hipaa-fine-tuning
|
|
patterns:
|
|
- pattern-either:
|
|
- pattern: $X.fine_tuning.$M(...)
|
|
- pattern: $X.fine_tuning.$Y.$M(...)
|
|
- pattern: $X.fine_tuning.$Y.$Z.$M(...)
|
|
message: |
|
|
OpenAI fine-tuning endpoints (`/v1/fine_tuning/*`) are NOT on the
|
|
HIPAA-eligible endpoints list. Customer data must not be uploaded
|
|
for fine-tuning under the BAA.
|
|
|
|
If you need a non-PHI fine-tuning job, suppress per-line with
|
|
`# nosemgrep: openai-non-hipaa-fine-tuning` and explain why.
|
|
languages: [python]
|
|
severity: ERROR
|
|
metadata:
|
|
category: compliance
|
|
subcategory: hipaa
|
|
confidence: HIGH
|
|
paths:
|
|
exclude:
|
|
- '**/.semgrep/**'
|
|
- '**/test/**'
|
|
- '**/tests/**'
|
|
- '**/test_*.py'
|
|
|
|
- id: openai-non-hipaa-realtime
|
|
patterns:
|
|
- pattern-either:
|
|
- pattern: $X.realtime.$M(...)
|
|
- pattern: $X.beta.realtime.$M(...)
|
|
- pattern: $X.beta.realtime.$Y.$M(...)
|
|
message: |
|
|
OpenAI Realtime API (`/v1/realtime`) is NOT on the HIPAA-eligible
|
|
endpoints list. Use chat completions or the responses API instead.
|
|
|
|
For non-PHI realtime usage, suppress per-line with
|
|
`# nosemgrep: openai-non-hipaa-realtime` and explain why.
|
|
languages: [python]
|
|
severity: ERROR
|
|
metadata:
|
|
category: compliance
|
|
subcategory: hipaa
|
|
confidence: HIGH
|
|
paths:
|
|
exclude:
|
|
- '**/.semgrep/**'
|
|
- '**/test/**'
|
|
- '**/tests/**'
|
|
- '**/test_*.py'
|
|
|
|
- id: openai-non-hipaa-legacy-completions
|
|
patterns:
|
|
- pattern: $X.completions.create(...)
|
|
- pattern-not: $Y.chat.completions.create(...)
|
|
message: |
|
|
Legacy text completions (`/v1/completions`) is NOT on the
|
|
HIPAA-eligible endpoints list. Use chat completions
|
|
(`client.chat.completions.create(...)`) or the responses API
|
|
(`client.responses.create(...)`) instead.
|
|
|
|
The legacy endpoint is also a deprecated surface — there is rarely
|
|
a good reason to use it. If you really need it for a non-PHI flow,
|
|
suppress per-line with `# nosemgrep: openai-non-hipaa-legacy-completions`.
|
|
languages: [python]
|
|
severity: ERROR
|
|
metadata:
|
|
category: compliance
|
|
subcategory: hipaa
|
|
confidence: HIGH
|
|
paths:
|
|
exclude:
|
|
- '**/.semgrep/**'
|
|
- '**/test/**'
|
|
- '**/tests/**'
|
|
- '**/test_*.py'
|