Skip to content

About

Offline Node.js evidence checker for CVE-2026-106074: Crypto entropy cache duplication in snapshots. 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-2026-106074-check by CompleteTech LLC

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

node-cve-2026-106074-check

Offline version-evidence checker for CVE-2026-106074 — Node.js: Crypto entropy cache duplication in snapshots.

It compares the Node.js versions you report against the fixed releases named in the official October 2026 Node.js security release and classifies each one as potentially_affected, fixed_version_reported, unsupported_runtime or unknown. Results are triage evidence, not a runtime assessment, a patch attestation or an exploit verification.

Contents

Scope

  • Reported versions only. The checker never scans the current machine. Supply the exact version taken from the deployed process or container; a public URL is not an inventory.
  • Offline. It never contacts websites, executes Node.js or project code, or sends exploit payloads, and no network mode is implemented.
  • Conservative. Hidden versions, custom builds, conflicting evidence and uncovered release lines remain unknown.
  • Exploitability is not tested. Every result carries exploitability: not_tested.
  • Authorized use. Only assess runtimes and inventories you are authorized to inspect.

Requirements

Python 3.10 or newer and the standard library only. There are no runtime dependencies, and node_cve_2026_106074_check.py can be copied and used on its own. CI runs the tests on Ubuntu and Windows with Python 3.10 and 3.14.

Quick start

# Assess one reported version; print JSON to stdout
python node_cve_2026_106074_check.py --version 22.23.4 --format json

# Assess the synthetic example inventory and write an HTML report
python node_cve_2026_106074_check.py --inventory examples/inventory.example.json --format html --output report.html

# Assess a CSV list; exit 1 if any result is potentially_affected or unsupported_runtime
python node_cve_2026_106074_check.py --csv-list inventory.csv --fail-on-finding

# List every option
python node_cve_2026_106074_check.py --help

CSV is the default format and stdout is the default destination. --version is repeatable and may be combined with --inventory and --csv-list; each value or entry describes a separate runtime. JSON, CSV and HTML are rendered from the same evidence. See Command-line options for details.

Synthetic report preview

Node.js synthetic report

Open the synthetic example report. It uses example data only and is not a live assessment.

Fixed versions and classification

Rules reviewed on 2026-10-24 against the official release advisory and the linked fixed releases (the advisory was published 2026-10-21).

Node.js major Fixed version
22.x 22.23.4
24.x 24.22.1
26.x 26.12.1

For each listed major, versions below its fixed version require remediation. 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.

Reported version Result
Plain version below the fixed version of a listed major potentially_affected
Fixed version, or a later plain version on the same listed major fixed_version_reported
Major that is end-of-life in the Node.js release schedule unsupported_runtime
Any other major (not listed and not end-of-life in the reviewed schedule) unknown
Missing, conflicting, malformed, prerelease, nightly or other qualified version unknown
Vendor, distribution, custom or unrecognized build source unknown

Notes on these rules:

  • unsupported_runtime is upgrade guidance, not a claim that this particular flaw exists in every historical build. It applies to majors 4–21, 23 and 25 as listed in the Node.js release schedule reviewed on 2026-10-24, and to major 0 by inference (the schedule lists only 0.8, 0.10 and 0.12, all ended by 2016-12-31).
  • Vendor, backported and custom builds need vendor evidence. A build source of unknown (the default), upstream, official or official_upstream, or none at all, is compared as an upstream release; any other value is treated as a vendor or custom build.
  • Unlisted, custom/backported and prerelease builds need explicit evidence. Do not label unknown builds safe.
  • fixed_version_reported is a statement about the reported version. It is not proof of the deployed runtime and not a verified deployed patch; a lockfile or patched workstation does not establish what the running process uses.
  • Versions are normalized by stripping ASCII whitespace and at most one leading v. Only plain MAJOR.MINOR.PATCH versions are compared.

Inputs

Each input describes one Node.js runtime. Inputs are limited as described in Limits, exit codes and interruption.

Field Required Description
schema_version JSON only Integer 1.
node_version Yes Version string reported for the runtime.
runtime No JSON only. If present, must be "node".
label No Name shown in reports. Defaults to the file name and entry number, such as inventory:1.
build_source No Provenance of the build; see Fixed versions and classification. Defaults to unknown.
startup_snapshot_or_sea_use_snapshot No Operator observation: startup snapshot or SEA use snapshot.
entropy_api_called_during_snapshot_build No Operator observation: entropy API called during snapshot build.

Observations are operator-supplied and unverified. They are recorded in the report as reported_true, reported_false or not_reported; they never change the version result and never establish reachability.

JSON inventory (--inventory) is one object or an array of objects. Observation values are booleans or null. Unknown keys, duplicate keys, non-finite numbers and invalid types are rejected. See examples/inventory.example.json.

CSV list (--csv-list) needs a header row with a node_version column. Optional columns are label, build_source and the two observation columns, whose cells must be exactly true, false or blank. Unknown or duplicate columns, rows with a different number of fields than the header, bare quotes inside unquoted cells, text after a closing quote and NUL characters are rejected. schema_version and runtime are not CSV columns. For example (illustrative values):

label,node_version,build_source,startup_snapshot_or_sea_use_snapshot,entropy_api_called_during_snapshot_build
api-1,24.22.1,,,
worker-2,22.23.3,upstream,true,false

Command-line options

At least one of --version, --inventory or --csv-list is required. They can be combined in one run.

Option Description
--version VERSION Assess an explicitly supplied Node.js version (a leading v is accepted). Repeatable. --node-version is an alias.
--build-source SOURCE Build provenance applied to all --version inputs, at most 512 characters. Default unknown, which is compared as upstream.
--startup-snapshot-or-sea-use-snapshot {true,false} Observation applied to all --version inputs.
--entropy-api-called-during-snapshot-build {true,false} Observation applied to all --version inputs.
--inventory FILE JSON object or array of entries. Repeatable.
--csv-list FILE CSV list. Repeatable.
--format {csv,json,html} Output format. Default csv.
--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, an existing file is never overwritten.
--brand NAME Organization name shown in the HTML report, at most 512 characters. Default CompleteTech LLC.
--fail-on-finding Exit 1 when any result is potentially_affected or unsupported_runtime.
--tool-version Print the tool name and version. Not the same as --version, which supplies a Node.js version.

--output must be a regular-file path that differs from every input file, even with --force. The --brand option affects only the HTML report.

Reports and results

Result statuses

status Meaning
potentially_affected The reported version is below the fixed version of its major and needs remediation review.
fixed_version_reported The reported version is at or above the fixed version. This is not deployed-patch attestation.
unsupported_runtime The major is end-of-life. Treat as upgrade guidance, not as confirmation of this flaw.
unknown Evidence is missing, conflicting, qualified, vendor-built or outside the rules. Unknown is not safe.

A more specific version_status accompanies each result:

status version_status values
potentially_affected below_fixed_version
fixed_version_reported at_or_above_fixed_version
unsupported_runtime unsupported_release_line
unknown version_not_disclosed, conflicting_versions, unrecognized_or_partial_version, qualified_version_unverified, vendor_build_unverified, release_line_not_in_rules

Every result also has exploitability: not_tested, evidence_type: operator_reported_version, and an evidence_confidence of low (or insufficient when the status is unknown). Operational failures such as unreadable input or an existing output file are not assessment results; they exit with status 2. Input validation failures occur before publication; output or cleanup failures can happen after bytes are written, so inspect any output before retrying. The notes and recommendation fields explain each verdict and the next step.

Formats

  • JSON has the top-level fields tool, tool_version, report_schema_version, cve, rules_as_of, sources, complete, generated_at and results. Each result carries input_label, input_source (version, inventory or csv), node_version, build_source, status, version_status, fixed_version, exploitability, evidence_type, evidence_confidence, cve, rules_as_of, one precondition_<observation key> field per observation, recommendation, notes, sources and report_schema_version. The report schema version is 1.0.
  • CSV has the same result fields as columns, plus record_type, report_complete, tool, tool_version and generated_at; list values are joined with |. Every field is quoted, and values that start with =, +, - or @ after any leading whitespace, or with a tab or line break, are prefixed with an apostrophe to neutralize spreadsheet formulas. Completed assessments use record_type=result. When no result exists (for example, an interruption before the first assessment), a single record_type=summary row preserves report_complete and the rule metadata without an asset or verdict.
  • HTML is a self-contained, script-free page with a restrictive Content-Security-Policy. It lists results (potentially affected first, fixed version reported last), the upstream fixed releases and the sources, and marks interrupted runs.

generated_at is a UTC ISO-8601 timestamp captured once per CLI run and reused across report formats and interruption recovery. It is distinct from the rule date (rules_as_of) and is not an evidence-acquisition time. Observation fields never change a verdict.

Limits, exit codes and interruption

Input limits. Regular UTF-8 files only. Per run: 4 MiB per file, 100 files, 10,000 entries in total (including --version values) and 512 characters per text field. Use stable local copies; file identity checks are not a sandbox against concurrent adversarial mutation.

Output. Reports are rendered and UTF-8 encoded in memory before anything is written, then published atomically to files. Without --force, an existing output file is never overwritten; if the filesystem cannot create hard links, the checker stops with an error and asks for another filesystem or an explicit --force. Output to stdout is not atomic. CSV output rejects NUL characters.

Exit codes.

Code Meaning
0 The run completed.
1 --fail-on-finding was set and at least one result is potentially_affected or unsupported_runtime.
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 The run was interrupted.

Interruption.

  • Before the output path is validated, nothing is written.
  • If a complete report has already been published, it is preserved and the checker exits 130 without retrying.
  • If file publication was started but its outcome is indeterminate, the checker exits 130 without retrying. Inspect the output before retrying; the diagnostic does not establish whether publication succeeded.
  • During stdout output, the document can be left partial, and a second report is never appended.
  • Otherwise the completed results are written with complete: false (report_complete=false in CSV), and the HTML report states that the run was interrupted. Assess the remaining inputs again.

Sources and provenance

Primary sources:

The embedded version engine is copied from node-cve-2026-105481-check commit de3161c4a8a7f7ca5c9e531df43b94e2805b52fd; no companion repository is required at runtime. Version metadata is not a patch or build signature.

Project documents

Keep real inventories, targets, reports and secrets outside version control; use synthetic examples in tests and documentation.

Contributing, security and license

Contributions are focused on this one CVE. Open a focused issue and pull request, use synthetic fixtures, and do not contact real targets in tests. Verify current primary sources before changing rules, and keep unknown evidence uncertain. Before submitting, run:

python -m unittest -v
python node_cve_2026_106074_check.py --help
git diff --check

Hosted CI must pass for the exact reviewed head. See CONTRIBUTING.md for contribution licensing and SECURITY.md for private vulnerability reporting.

Licensed under GNU AGPL-3.0 (LICENSE) or a separate commercial license from CompleteTech LLC; see LICENSING.md. CompleteTech LLC names, logos and the repository artwork in docs/images/ are brand assets that are not covered by the code license; see BRAND_ASSETS.md.

About

Offline Node.js evidence checker for CVE-2026-106074: Crypto entropy cache duplication in snapshots. 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