Skip to content
This repository was archived by the owner on Mar 20, 2026. It is now read-only.

Release/0.1.2#6

Merged
F0RLE merged 3 commits into
mainfrom
release/0.1.2
Mar 13, 2026
Merged

Release/0.1.2#6
F0RLE merged 3 commits into
mainfrom
release/0.1.2

Conversation

@F0RLE
Copy link
Copy Markdown
Owner

@F0RLE F0RLE commented Mar 13, 2026

Summary

Release v0.1.2 with Signal Quality improvements.

Features Implemented

  • Proxy geo-location validation (fetch_country)
  • Smart comparison pre-filter (~55% domain reduction)
  • Compare-Scan UI Polish
  • 429 Exponential Backoff with Retry-After
  • Deduplicate Body Readers (macro)
  • Confidence boost on network evidence
  • Configurable DOM Geo-markers (profiles.toml)
  • Adaptive Body Skip Threshold (Content-Type based)
  • Redirect chain deep scan infrastructure

Fixes

  • Variable shadowing in analysis.rs
  • local_confidence field in ComparisonResult
  • Updated tests

Roadmap

  • v0.1.2 completed, complex items moved to future releases

Copy link
Copy Markdown

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f2fc36b97d

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".

Comment thread src/scanner.rs
Comment on lines +782 to +786
.split('|')
.find_map(|part| {
let p = part.trim();
p.strip_prefix("retry-after=")
.and_then(|s| s.parse::<u64>().ok())
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Parse Retry-After from headers before scheduling 429 retries

scan_domain now derives backoff from attempts[0].reason by looking for retry-after=..., but that field is populated by analyze_http_observation as HTTP {status_code} {status_label} and does not include response headers, so this parser always falls back to 0 and then clamps to a ~1s wait. For domains that send 429 with a real Retry-After (e.g. 10–60s), the scanner will still hammer retries almost immediately, which both defeats the intended throttling behavior and can skew classification toward persistent rate-limit failures.

Useful? React with 👍 / 👎.

Comment thread src/scanner/analysis.rs Outdated
Comment on lines +500 to +504
if let Some(mut redirect_signal) = classify_redirect(final_url) {
redirect_signal.reason = format!("Final URL: {}", redirect_signal.reason);
redirect_signal.confidence = redirect_signal.confidence.saturating_add(5).min(95);
choose_better_signal(&mut best_signal, redirect_signal.clone());
signals.push(redirect_signal);
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Avoid counting final URL redirect evidence twice

analyze_http_observation already records classify_redirect(final_url) before body analysis, but this new block invokes it again and pushes a second (boosted) signal from the same URL. Because should_block treats verdict_consensus >= 2 as corroboration, a single redirect hint can now satisfy the multi-signal rule by itself, causing false blocked classifications on otherwise successful responses (for example, benign URLs containing marker-like path fragments).

Useful? React with 👍 / 👎.

@F0RLE F0RLE merged commit 0acc28e into main Mar 13, 2026
3 checks passed
@F0RLE F0RLE deleted the release/0.1.2 branch March 13, 2026 18:53
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant