Repository navigation
fix: keep validators usable where O_NOFOLLOW is unavailable (Windows) - #113
Open
PhucNguyen-Dev wants to merge 1 commit into
Open
PhucNguyen-Dev wants to merge 1 commit into
PhucNguyen-Dev wants to merge 1 commit into
Conversation
Windows Node leaves fs.constants.O_NOFOLLOW and O_NONBLOCK undefined, so readFileWithinLimit() in both validators threw "OS no-follow and nonblocking input protection is unavailable" for every file. On Windows the findings validator therefore failed 7 of its CLI tests and the coverage-ledger validator silently skipped its CLI tests instead. Fall back to an lstat symlink check and drop the unavailable flags from the open call when the constants are absent; the ELOOP/EMLINK handling and every other safeguard stay unchanged. The test suites no longer gate their CLI cases on the constants, so both run on every platform. Windows: validate-findings 22 pass/7 fail -> 32 pass/0 fail, validate-coverage-ledger 24 pass/7 skipped -> 29 pass/0 fail (2 skips remain: FIFO cases need Git-Bash, symlink creation needs Developer Mode). Linux/macOS is unaffected: the constants exist there, so the original open path is taken unchanged.
agammann
reviewed
Nov 3, 2026
agammann
left a comment
There was a problem hiding this comment.
In both validators, lstatSync(file) followed by openSync(file, O_RDONLY) is not an atomic replacement for O_NOFOLLOW: another process can replace the file with a symlink between those calls, and the open will follow it. This changes the previous fail-closed behavior precisely on platforms without O_NOFOLLOW. Also, when O_NONBLOCK is absent, opening a FIFO can block before the later fstat rejects it. Could these platforms keep a fail-closed path until a race-safe open strategy exists, with tests for the missing-flag cases?
coffeegrind123
added a commit
to coffeegrind123/security-audit-skill
that referenced
this pull request
Nov 11, 2026
Upstream main has not moved since 2026-09-14, but 79 pull requests are open. Each was read at the diff level, tested in its own worktree, and checked for conflicts; the best of each duplicate group is ported. Validators: cloudflare#143 archived verifier owner reuse, cloudflare#87 shared artifacts within one attempt, cloudflare#102/cloudflare#103 duplicate JSON keys, cloudflare#111 1,000-field objects, cloudflare#127 ordinary class must match assignment, cloudflare#104 self-excluded ordinary block, cloudflare#141 verbatim quotes checked with --target, cloudflare#45 line upper bound, cloudflare#84 --terminal rejects open units, cloudflare#145 identity-checked open on Windows (POSIX still fails closed), cloudflare#48 retained-artifact check (skills/ part only). Overlapping changes were merged so both behaviours and both test sets survive. Method: cloudflare#125 critic reasons through handoff, cloudflare#132 deployment controls become units, cloudflare#83 carried-record gates, cloudflare#93 abort cleanup, cloudflare#23 streaming tool calls, cloudflare#50 config-dependent reachability stays needs_validation, cloudflare#58 sequential verifier fallback, cloudflare#96 verifier no longer sees the hunter's verdict, severity or fix. Execution safety (cloudflare#76 neutralized git, symlink-safe copies; cloudflare#95 sanitized repo ID and prior-run compatibility; cloudflare#108 untrusted-content prompt block; cloudflare#129 isolation built beside the host) moves into EXECUTION-SAFETY.md so SKILL.md stays under its token budget. Companions: cloudflare#101 cryptography, cloudflare#116 Android, cloudflare#35 edge isolate state. Skipped: duplicates (cloudflare#82 cloudflare#106 cloudflare#91 cloudflare#70 cloudflare#138 cloudflare#131 cloudflare#75 cloudflare#63 cloudflare#97 cloudflare#123 cloudflare#25 cloudflare#26 cloudflare#16 cloudflare#40 cloudflare#42 cloudflare#142), validator-weakening fallbacks (cloudflare#59 cloudflare#113 cloudflare#99), README/LICENSE churn, rebrands, scope creep, and cloudflare#85 cloudflare#90 cloudflare#8 cloudflare#89 cloudflare#32. UPGRADE.md records this sync procedure. Co-authored-by: Rohankollabathula67 <rohankollabathula67@gmail.com> Co-authored-by: lab 1207 <hajarevaibhav527@gmail.com> Co-authored-by: Rudy Celekli <rudy@gradiahq.com> Co-authored-by: Mohit Rawat <rawatmohit17906@gmail.com> Co-authored-by: MAHER AL-AQIL <18458259+maqalaqil@users.noreply.github.com> Co-authored-by: CodeInf <jiuqiaoy@gmail.com> Co-authored-by: pucedoteth <119044801+pucedoteth@users.noreply.github.com> Co-authored-by: Matt Van Horn <mvanhorn@users.noreply.github.com> Co-authored-by: Pizo <pizofreude@proton.me> Co-authored-by: ZhiHe-ma <302089060+ZhiHe-ma@users.noreply.github.com> Co-authored-by: shivam <juyaljii610@gmail.com> Co-authored-by: Dante <duanjl.china@gmail.com> Co-authored-by: Francisco <francisco@theysay.me> Co-authored-by: Roman Dolgov <alter@users.noreply.github.com> Co-authored-by: bodapatisaikrishna <saikrishnabodapati@gmail.com> Co-authored-by: alvessh <117163985+alvessh@users.noreply.github.com> Co-authored-by: itinance <hhuebel@itinance.com> Co-authored-by: ChoYugyeong <cho4639@gmail.com> Co-authored-by: Shaun Liebenberg <shaun@liebies.co.za> Co-authored-by: JAI <jai19kharb@gmail.com> Co-authored-by: ghostzeronz-coder <ghostzeronz-coder@users.noreply.github.com> Co-authored-by: Aman Mohamed Ali <69119173+spidervirus@users.noreply.github.com> Co-authored-by: Sandeep.dhami <sandeep84397@gmail.com> Co-authored-by: jastfan <duohello640@gmail.com> Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This was referenced Nov 18, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
readFileWithinLimit()in both validators refuses to run at all whenfs.constants.O_NOFOLLOW/O_NONBLOCKare unavailable:Windows Node leaves both constants
undefined, so on Windows every CLI invocation dies with that message. Phase 4 and Phase 5 validation — the whole point offindings.jsonandcoverage-ledger.jsonbeing machine-checked — is unusable there. The bundled suites show it: the findings suite fails 22 pass / 7 fail, and the ledger suite hides the same bug by skipping its CLI tests (HAS_SAFE_INPUT_OPEN).Change
lstatbefore opening and drop the flags from theopencall.ELOOP/EMLINKhandling, sameSafeInputErrormessages.POSIX behaviour is unchanged — the constants exist there, so the original
openSync(..., O_RDONLY | O_NOFOLLOW | O_NONBLOCK)path is taken exactly as before.Test results
Windows 11, Node v22.23.4, from
skills/security-audit:validate-findings.test.cjsvalidate-coverage-ledger.test.cjsThe 2 remaining skips are platform-inherent, not regressions: FIFO cases need Git-Bash (the repo's existing
win32guard stays), and file-symlink creation on Windows needs Developer Mode. Both are unchanged by this patch.Honest caveat
lstat-then-openis not as race-proof asO_NOFOLLOWon its own — it is a check followed by an open rather than one atomic flag. On platforms where the flag exists nothing changes, so the weaker path only applies where today the alternative is a hard failure. If you would rather keep the atomic guarantee and only accept Windows input via stdin, say so and I will rework it.Each validator still rejects a symlink when handed one from a platform that can create it; that is the behaviour the previously platform-skipped test asserts.