Skip to content

chore(deps): Update openai requirement from <4.0.0,>=1.40.0 to >=3.13.0,<4.0.0 #384

chore(deps): Update openai requirement from <4.0.0,>=1.40.0 to >=3.13.0,<4.0.0

chore(deps): Update openai requirement from <4.0.0,>=1.40.0 to >=3.13.0,<4.0.0 #384

Workflow file for this run

name: CodeQL
# Static analysis (SAST) for the Python engine. Satisfies the OpenSSF
# Scorecard SAST check and — more to the point — runs real code scanning on
# every push/PR so a regression surfaces in the Security tab rather than in
# production. Default query suite (high-signal security queries) to keep the
# baseline triage-able; promote to `security-extended` once it is triaged.
#
# Scope note: this pass scans `python` only. The JavaScript verifier surface
# (single_file_demo / standalone_verifier / worker) is security-critical and a
# good follow-up to add here behind a `paths-ignore` for the vendored bundle —
# kept out of this first pass to bound new-alert noise.
#
# All third-party Actions are SHA-pinned (R0.3 / scripts/lint_workflow_pins.py),
# reusing the same commit SHAs already vetted in the other workflows.
on:
push:
branches: [ "main" ]
pull_request:
branches: [ "main" ]
schedule:
- cron: '27 4 * * 1' # Weekly, Mondays 04:27 UTC (off-peak).
permissions:
contents: read
jobs:
analyze:
name: Analyze (python)
runs-on: ubuntu-latest
permissions:
security-events: write # upload CodeQL results to code scanning
actions: read
contents: read
steps:
- name: Harden runner (audit-only)
uses: step-security/harden-runner@e14015d583714f6e62063499dc959a02595150a1 # v2.21.1
with:
egress-policy: audit
- name: Checkout
uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- name: Initialize CodeQL
uses: github/codeql-action/init@cdf488f595d80d6e07e03d4674febd5ab45fa938 # v4.37.9
with:
languages: python
build-mode: none
- name: Perform CodeQL Analysis
uses: github/codeql-action/analyze@cdf488f595d80d6e07e03d4674febd5ab45fa938 # v4.37.9
with:
category: "/language:python"