TaskAttest is pre-1.0. Security fixes are applied to the latest tagged release. Older pre-1.0 releases are unsupported after a newer release is available.

Use GitHub private vulnerability reporting. Do not open a public issue for command execution, path traversal, environment leakage, process cleanup, resource-limit bypass, no-clobber, source-identity, log-store, or receipt-integrity vulnerabilities.

Include the TaskAttest version, operating system, repository shape with private content removed, command, structured error or receipt, and a minimal synthetic reproduction. Do not attach real logs: complete check logs may contain secrets.

An acknowledgement is targeted within 7 days. The maintainer will validate the report, coordinate disclosure, add a regression test, and publish a GitHub Security Advisory when appropriate. Targets are goals, not a service-level agreement for this volunteer project.

• TaskAttest executes repository-defined code. Use it only for repositories whose code and configuration you are prepared to run.
• TaskAttest spawns argument arrays and never interpolates a shell command. npm, pnpm, Yarn, and Bun may themselves invoke package scripts through a platform shell.
• Environment inheritance is minimized, but explicitly forwarded variables are available to child processes.
• Runtime and log budgets are enforced, and Unix process groups or Windows process-tree termination are used on cancellation. They are resource controls, not an operating-system sandbox.
• Network and filesystem access are not sandboxed.
• Full log blobs are unredacted and potentially_sensitive. Summary redaction is best-effort and must not be treated as a data-loss-prevention control.
• Receipt digests detect mutation. They do not authenticate an author, attest a reproducible environment, or prove that discovery was sufficient.

Dependabot monitors Rust and GitHub Actions dependencies. CI checks Cargo.lock against RustSec advisories. Tagged releases are built only by the release workflow and include checksums, SBOMs, and GitHub/Sigstore attestations. See RELEASING.md.

Pull requests are checked with GitHub Dependency Review and fail when they introduce a dependency with a known moderate-or-higher-severity vulnerability. A weekly OpenSSF Scorecard analysis publishes authenticated results and uploads SARIF findings to GitHub code scanning. CodeQL default setup analyzes Rust and workflow sources with extended security queries. ClusterFuzzLite runs the production strict receipt parser on every code-changing pull request and in a longer weekly AddressSanitizer batch; see FUZZING.md.