Repository navigation
feat(repo,cli): close the four P32 TLS follow-ons - #208
Conversation
- sc serve fingerprint and --tls auto-mint refuse a root with no .sc/. - Warn when an SC_HTTPS_* knob is set but the URL is plaintext sc+http://. - Bound the sc+http(s) client's connect, TLS handshake, opening and wire handshake (30 s) so a silent peer fails the connect; lift it afterwards. - sc remote add parse-validates sc+http(s):// URLs at add time. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
Note Currently processing new changes in this PR. This may take a few minutes, please wait... ⚙️ Run configuration
📒 Files selected for processing: 10
✨ Finishing Touches
Comment |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 54d053bdee
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
… discover the repo Codex: a timeout in the TLS handshake or wire handshake now surfaces as ConnectionLost like the opening read; the connect shares one deadline across resolved addresses; an empty CA knob counts as set for the downgrade warning; sc serve fingerprint finds the repo from a subdirectory; and the reserved sc+http:/sc+https: prefix, not just the exact form, is parse-validated at add. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Reviewer: the lift test passed with the lift removed, so it now inspects the live socket options (and fails without the lift). A bucket-backed server opens its bucket before answering HELLO, so the wire handshake gets 300 s while the connect, TLS and opening keep 30 s. Docs say these are inactivity bounds. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 987f6bab8f
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
…knobs Codex: to_socket_addrs ran before the deadline could apply, so a stalled resolver defeated the connect bound; and var() hid a non-UTF-8 SC_HTTPS_* value from the downgrade warning. Resolve on a helper thread under the deadline and read the knobs with var_os. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1b1fe9a1b7
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Codex: write_client_opening turned every write failure into InvalidArgument, so a peer that never reads (opening larger than the socket buffers) hit the new write bound but got the wrong typed error. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: ab830f18cd
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Codex: with one shared deadline, a first address that silently drops packets could use all of it and starve a reachable later address. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
tonibergholm-codento
left a comment
There was a problem hiding this comment.
Required checks green on 681d5ea; reviewer and Codex findings fixed and resolved.
Closes the P32 follow-ons item in ROADMAP.md.
sc serve fingerprintand--tlsauto-mint refuse a root with no.sc/, namingsc init, instead of minting a stray identity.SC_HTTPS_STRICT/FINGERPRINT/KNOWN_HOSTS/CA_FILE/CAwith a plaintextsc+http://URL prints a stderr warning (the knob does nothing there).ConnectionLost) instead of hanging it; the bound is lifted after the handshake so long transfers are unaffected. Covered for plaintext and TLS, plus a test that an idle connection outlives the bound.sc remote addparse-validatessc+http(s)://URLs at add time, likessh://and the bucket schemes.The client bound is not configurable. ADR-0042 and ROADMAP.md updated.
cargo clippy --workspace --all-targets -- -D warningsclean;cargo test --workspace993 passed, 0 failed.🤖 Generated with Claude Code
Summary by CodeRabbit
sc remote addnow rejects malformedsc+http://andsc+https://URLs before saving them.sc+http://URL now triggers a warning.