Security¶
Supply chain security¶
repomatic implements most of the practices described in Astral’s Open Source Security at Astral post, baked into a drop-in setup that any maintainer can inherit by pointing their workflows at the reusable callers.
Astral practice |
How |
|---|---|
Ban dangerous triggers ( |
The lint-workflow-security job runs |
Minimal workflow permissions |
|
Pinned actions |
All |
No force-pushes to |
|
Immutable release tags |
|
Dependency cooldowns |
Renovate stabilization windows ( |
Trusted Publishing |
PyPI uploads via OIDC with no long-lived token. The |
Cryptographic attestations |
Every binary and wheel is attested to the workflow run that built it via |
Checksums in installer scripts |
The |
Fork PR approval policy |
|
Warning
Known gap: multi-person release approval. Astral gates releases behind a dedicated GitHub deployment environment with required reviewers, so that a single compromised account cannot publish. repomatic does not enforce this, but if the repository has multiple maintainers, I recommend adding an environment: release key to the caller-side publish-pypi job (and to the upstream create-release job, if the caller exposes it) in the downstream workflow and configuring required reviewers on that environment in repo settings.
Important
One-time PyPI Trusted Publisher setup. Each downstream repository must register a Trusted Publisher entry on PyPI for its own caller workflow. The publisher config matches against the OIDC job_workflow_ref claim, which names the downstream’s workflow file (typically .github/workflows/release.yaml). Without this registration, the first PyPI upload after migration fails cleanly with a publisher mismatch error. See the PyPI Trusted Publishers documentation for the registration steps.
Third-party action minimization¶
Every third-party GitHub Action executes with access to GITHUB_TOKEN and repository secrets. Each action is a trust delegation: you depend on the maintainer’s security practices, their CI pipeline, and their transitive dependencies. A compromised action can steal secrets, inject code into builds, or tamper with releases.
repomatic has systematically eliminated 18 third-party actions since late 2025, replacing them with internal CLI commands, SHA-256-verified binary downloads, and runner built-in tools:
Removed action |
Replacement |
Strategy |
|---|---|---|
|
|
Internal CLI |
|
|
Internal CLI |
|
|
Internal CLI |
|
|
Internal CLI |
|
|
Direct binary + SHA-256 |
|
|
Direct binary + SHA-256 |
|
|
Direct binary + SHA-256 |
|
|
Direct binary + SHA-256 |
|
|
Direct binary + SHA-256 |
|
Direct |
Direct binary + SHA-256 |
|
|
Runner built-in |
|
Bash + |
Runner built-in |
|
Runner built-in Rust |
Runner built-in |
|
|
Consolidated |
|
|
First-party replacement |
|
|
Pinned CLI |
|
|
Pinned CLI |
|
Explicit |
Removed entirely |
The remaining third-party actions (5 of 14 total) are:
Action |
Purpose |
|---|---|
|
Core toolchain: installs |
|
Creates autofix PRs |
|
Locks inactive issues |
|
Dependency updates |
|
Debug diagnostics (no secrets access) |
Replacement strategies, ordered from most to least isolated:
Internal CLI: the operation runs inside
repomaticPython code with no external process.Direct binary download: checksummed binary fetched from a GitHub release URL, no action code path involved.
Runner built-in: uses tools pre-installed on the GitHub Actions runner (
gh, Rust toolchain).First-party replacement: swaps a community action for an official
actions/*equivalent maintained by GitHub.
Ruff consolidation¶
Eight separate Python linters and formatters have been absorbed into ruff, eliminating eight runtime or dev dependencies:
Removed tool |
What it did |
Replaced |
|---|---|---|
|
Static analysis and linting |
Feb 2023 |
|
Docstring convention enforcement |
Feb 2023 |
|
Unused import removal |
Feb 2023 |
|
Python syntax modernization |
Feb 2023 |
|
Import sorting |
Feb 2023 |
|
Code formatting |
Sep 2023 |
|
Docstring formatting |
Jan 2024 |
|
Python formatting in Markdown code blocks |
Feb 2026 |
The mdformat-black plugin was also swapped for mdformat-ruff (Aug 2024): same dependency count, but aligns the Markdown pipeline with ruff’s formatting rules.
autopep8 is the only legacy formatter still in use: it handles long-line comment wrapping that ruff does not yet cover.
uv consolidation¶
Five separate packaging and install tools have been absorbed into uv, which now handles dependency management, builds, publishing, auditing, and Python version installation:
Removed tool |
What it did |
Replaced |
|---|---|---|
|
Dependency management, lock files, virtual environments |
Jun 2024 |
|
Package building (wheels and sdists) |
Sep 2024 |
|
PyPI uploads |
Jan 2025 |
|
Wheel validation |
Jan 2025 |
|
Vulnerability scanning |
Mar 2026 |
uv also consolidated command-line usage that previously required separate tools: pip install became uv pip install / uv sync, pipx became uvx, and actions/setup-python was replaced by astral-sh/setup-uv (counted in the action minimization table above).
Two other Python packages were eliminated outside the ruff/uv consolidations: pipdeptree (replaced by an internal deps-graph implementation) and gitignore-parser (replaced by py-walk).
Permissions and token¶
Several workflows need a REPOMATIC_PAT secret to create PRs that modify files in .github/workflows/ and to trigger downstream workflows. Without it, those jobs silently fall back to the default GITHUB_TOKEN, which lacks the required permissions.
After your first push, the setup-guide job automatically opens an issue with step-by-step instructions to create and configure the token.
Concurrency and cancellation¶
All workflows use a concurrency directive to prevent redundant runs and save CI resources. When a new commit is pushed, any in-progress workflow runs for the same branch or PR are automatically cancelled.
Workflows are grouped by:
Pull requests:
{workflow-name}-{pr-number}— Multiple commits to the same PR cancel previous runsBranch pushes:
{workflow-name}-{branch-ref}— Multiple pushes to the same branch cancel previous runs
release.yaml uses a stronger protection: release commits get a unique concurrency group based on the commit SHA, so they can never be cancelled. This ensures tagging, PyPI publishing, and GitHub release creation complete successfully.
Additionally, cancel-runs.yaml actively cancels in-progress and queued runs when a PR is closed. This complements passive concurrency groups, which only trigger cancellation when a new run enters the same group — closing a PR doesn’t produce such an event.
Tip
For implementation details on how concurrency groups are computed and why release.yaml needs special handling, see the repomatic.github.actions module docstring.
AV false-positive submissions¶
Compiled Python binaries (built with Nuitka --onefile) are frequently flagged as malicious by heuristic AV engines. The onefile packaging technique (self-extracting archive with embedded Python runtime) triggers generic “packed/suspicious” signatures. This is a known issue across the Nuitka ecosystem.
The scan-virustotal job in release.yaml uploads all compiled binaries to VirusTotal on every release. This seeds AV vendor databases to reduce false positive rates for downstream distributors (Chocolatey, Scoop, etc.).
When a release is flagged, the /av-false-positive skill generates per-vendor submission files with pre-written text and form field mappings. The vendor details below document the process for manual reference.
Vendor portals¶
Vendor |
Engines covered |
Portal |
Format |
Turnaround |
|---|---|---|---|---|
Microsoft |
|
One file per form, 1900 char limit on additional info |
Fastest |
|
BitDefender |
|
One file per form, screenshot mandatory |
Fast |
|
ESET |
|
Email to |
Single email, password-protected ZIP ( |
Reliable |
Symantec |
|
Hash submission only (no |
3-7 business days |
|
Avast/AVG |
|
One file per form, shared engine |
Medium |
|
Sophos |
|
One file per form, 25 MB max per submission |
Up to 15 business days |
Submission priority¶
Submit in this order to maximize impact:
Microsoft: most influential engine. ML detections (
Sabsik,Wacatac) have the broadest downstream effect.BitDefender: powers ~6 downstream vendor engines. Highest detection-removal-per-submission ratio.
ESET: email-based channel with no portal dependency. The most reliable submission path.
Symantec: ML detections (
ML.Attribute.*) may take longer to process.Avast/AVG: shared engine, so one submission covers both.
Sophos: PUA detections require justification of the software’s legitimate purpose.
Submission content¶
Every false-positive submission should include:
The binary’s VirusTotal report link.
VirusTotal links for the clean
.whland.tar.gzsource distributions (as comparison evidence).The GitHub release link and direct download URL for the binary.
Project homepage and PyPI URL.
License from
pyproject.toml.Reference to any prior false-positive issue in the repository.
All submission text should mention that the binary is compiled with Nuitka --onefile from an open-source project.
Known portal issues¶
Microsoft: CORS errors or stuck progress modals during upload (auth session expiring). Workaround: sign out, clear cookies for
microsoft.com, sign back in, submit immediately.BitDefender: form sometimes returns “Your request could not be registered!” with no details. Retry later.
Avast: form sometimes returns “An internal error occurred while sending the form.” Retry later.