Skip to content

About

Offline Node.js evidence checker for CVE-2026-106071: Permission Model wildcard path over-grant. Single-file Python; JSON/CSV/HTML reports; no exploit payloads.

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

node-cve-2026-106071-check by CompleteTech LLC

Tests Python 3.10+ No runtime dependencies AGPL-3.0 or commercial license

node-cve-2026-106071-check

Offline evidence checker for CVE-2026-106071 — Node.js: Permission Model wildcard path over-grant. It compares operator-reported Node.js versions with the fixed release named in the vendor advisory and reports what that evidence does and does not support.

  • Offline and read-only. It never contacts websites, executes Node.js or project code, or sends exploit payloads. There is no network mode and no automatic scan of the current machine.
  • Evidence, not verification. Results describe the versions you report. Hidden versions, custom builds, conflicting evidence and uncovered release lines remain unknown. A public URL is not an inventory.
  • Exploitability is never tested. Every result carries exploitability: not_tested.
  • Standard library only. Python 3.10+ with no runtime dependencies; the single executable can be copied on its own.

Contents

Requirements and quick start

Requirements: Python 3.10 or newer. Nothing needs to be installed.

Supply an exact version taken from the deployed process (for example from the running container or service), not from a workstation or lockfile:

python node_cve_2026_106071_check.py --version 26.12.1 --format json

Assess a JSON inventory and write an HTML report, or fail a pipeline on findings:

python node_cve_2026_106071_check.py --inventory examples/inventory.example.json --format html --output report.html
python node_cve_2026_106071_check.py --csv-list inventory.csv --fail-on-finding

Show all options and run the local-only test suite:

python node_cve_2026_106071_check.py --help
python -m unittest -v

Synthetic report preview

Node.js synthetic report

Open the synthetic example report. Example data only; this is not a live assessment.

Advisory rules

Reviewed 2026-10-24 against the official release advisory and the linked fixed release. The rule snapshot date is reported as rules_as_of.

Node.js major Fixed version
26.x 26.12.1

How a plain, official-upstream version (an optional leading v is accepted) is classified:

Reported version Status
Below the fixed version on a listed major potentially_affected
The fixed version or a later plain version on the same listed major fixed_version_reported
A major past end-of-life in the Node.js release schedule unsupported_runtime
Any other major, or missing, malformed, prerelease, conflicting or vendor/custom/backported evidence unknown
  • The vendor does not specify an introduced version in the release advisory, so every earlier version of a listed major is treated as potentially_affected.
  • unsupported_runtime is upgrade guidance for end-of-life lines. It is not a claim that this particular flaw exists in every historical build. The end-of-life list comes from the Node.js release schedule, reviewed 2026-10-24.
  • Other active or future majors are not assumed to be patched; they remain unknown.
  • For each listed major, versions below its fixed version require remediation. Unlisted, custom/backported and prerelease builds need explicit evidence; do not label unknown builds safe.
  • A build_source other than blank, unknown, upstream, official or official_upstream (for example a distribution or vendor name) yields unknown. Vendor, backported and custom builds need vendor evidence.

Interpreting results

Status Meaning
potentially_affected The reported version needs remediation review. It is not confirmed exposure or exploitability.
fixed_version_reported The reported version is at or above the upstream fixed release. This is a reported version, not deployed-patch attestation or a safety guarantee.
unsupported_runtime The major is end-of-life; migrate to a supported, fixed release.
unknown Evidence is missing or unclassifiable. Missing metadata does not show a runtime is unaffected.

Each result also has a more specific version_status, evidence_type: operator_reported_version, and a low (or, for unknown, insufficient) evidence_confidence. Version metadata is not a patch or build signature.

Precondition observations

You may record three operator-supplied observations alongside a version. They are kept separate: they never change the version result and never establish reachability.

Observation Meaning
permission_model_enabled permission model enabled
wildcard_filesystem_grant_directly_under_root wildcard filesystem grant directly under root
shorter_target_path shorter target path

Each is true, false or not reported, and appears in reports as reported_true, reported_false or not_reported.

Usage

There is no automatic scan. Provide at least one of --version, --inventory or --csv-list.

Option Purpose
--version VERSION Assess an explicitly supplied version; repeatable, and each value describes a separate runtime. --node-version is an alias.
--build-source SOURCE Provenance of the --version builds (at most 512 characters): unknown (default, compared as upstream), upstream, official, or a vendor name.
--permission-model-enabled, --wildcard-filesystem-grant-directly-under-root, --shorter-target-path Each takes true or false and applies that observation to the --version inputs.
--inventory FILE JSON inventory; repeatable.
--csv-list FILE CSV inventory; repeatable.
--format csv|json|html Report format; default csv. All formats use the same evidence.
--output FILE Write the report to a new file instead of stdout. The directory must already exist.
--force Replace an existing output file. Without it, existing reports are protected. Input files can never be overwritten, even with --force.
--brand NAME Organization name shown in the HTML report (at most 512 characters).
--fail-on-finding Exit 1 when any result is potentially_affected or unsupported_runtime.
--tool-version Print the tool name and version.

Inputs

JSON inventory (--inventory): one object or an array. See examples/inventory.example.json.

  • Required: integer schema_version: 1 and string node_version.
  • Optional: runtime (only "node"), label, build_source, and the three observation keys above, each boolean or null.
  • Unknown keys, duplicate keys, non-finite numbers and invalid types are rejected.

CSV list (--csv-list):

  • Requires a header row with a node_version column.
  • Optional columns: label, build_source, and the observation columns (true, false, or blank).
  • Undocumented or duplicate columns, rows with a different number of fields than the header, and NUL characters are rejected.

Input files must be regular UTF-8 files. Use stable local copies: file identity checks are not a sandbox against concurrent adversarial mutation.

Output and reports

  • CSV quotes every field and neutralizes formula prefixes, including after leading whitespace. Completed assessments use record_type=result. When no result exists, a record_type=summary row preserves report_complete and rule metadata without an asset or verdict. NUL characters are rejected before a report is written.
  • JSON contains the tool and rule metadata, sources, complete, generated_at, and a results array.
  • HTML is a script-free report that shows the same run timestamp.

Every report carries canonical tool, tool_version and timezone-qualified UTC generated_at metadata. CSV appends these columns to each result and to an empty summary; JSON keeps the same fields. One timestamp is captured per CLI run and reused during interrupted-report recovery. It is report-generation time, not an invented observation or acquisition time, and it remains distinct from rules_as_of.

Reports are fully rendered in memory, then atomically published to files. Keep real inventories, targets, reports and secrets outside version control. Synthetic examples are in examples/ and docs/example-report.html.

Exit codes, interruption and limits

Exit code Meaning
0 The run completed.
1 At least one finding, only with --fail-on-finding.
2 Usage, input, output or cleanup failure (an operational error, not an assessment result). This does not establish whether report bytes were already written; inspect the output before retrying.
130 Interrupted.

Interruption behavior:

  • Completed rows are retained and written with the report marked incomplete (report_complete false in CSV, complete: false in JSON) when output is possible.
  • An interruption after a complete report is published preserves that report and exits 130.
  • If an interrupt leaves file publication indeterminate, the checker exits 130 without retrying. Inspect the output before retrying; the diagnostic does not establish whether publication succeeded.
  • An interruption during stdout output can leave a partial document and never appends a second report.

Limits: 4 MiB per file, 100 input files, 10,000 entries in total, and 512 characters per text field.

Sources

The embedded version engine is copied from node-cve-2026-105481-check commit de3161c4a8a7f7ca5c9e531df43b94e2805b52fd; no companion repository is required at runtime.

Documentation, contributing and license

About

Offline Node.js evidence checker for CVE-2026-106071: Permission Model wildcard path over-grant. Single-file Python; JSON/CSV/HTML reports; no exploit payloads.

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages