Releases
v1.11.0
Compare
Sorry, something went wrong.
No results found
1.11.0 (2026-11-02)
Features
agent-config and workflow-security rules, CI exit policy, repo-aware init, faster TS and diff scans (#192 ) (145eb0d )
ci: add GitHub Actions workflow-security rules (c082b50 )
cli: add -fail-on, -min-confidence and -quiet; gate the banner on TTY/CI/NO_COLOR (a9b149a )
cli: add "did you mean" suggestions and validate flags before scan output (cce068e )
cli: richer explain output and rules filters (7399a07 )
context: generate and refresh agent docs with codeguard context init|update (2a75286 )
doctor: probe govulncheck, check entrypoints and target languages (e3581ef )
init: detect the repository and write a small config; scan without one (3377568 )
MCP finding filters, did-you-mean errors, gitignore-aware walk, concurrent quality passes, security precision (#193 ) (9b001b4 )
mcp: summary-first, filtered, paginated findings for scan and validate_patch (668c56a )
prompts: govern Claude settings, MCP servers and hidden Unicode in agent config (a2243bb )
report: richer SARIF for GitHub code scanning (3a6896f )
rules: verified examples, confidence, and config keys in the catalog (bf88819 )
scan: honour .gitignore in the file walk (f71df06 )
security: honor justified gosec directives for shell-execution (f7f8423 )
security: show govulncheck stderr in failure diagnostics (fcf1880 )
shared taint engine, gocyclo-accurate complexity, richer SARIF, teaching explain, context generation, docs coverage (#194 ) (c6da761 )
Bug Fixes
action: make the sticky PR comment actually post (b2c9020 )
ai: remove the config-selectable mock triage provider (8979ac5 )
cache: keep security-secrets diagnostics on findings-cache hits (d7c3b9d )
cache: key per-file findings on the cross-file inputs they read (5bed0ad )
config: keep an explicit 0 for risk-scoring and production-risk weights (b04758d )
config: report every validation failure, not just the first (c59bb5b )
mcp: keep the 1.10 full report as the scan/validate_patch default (719bf8f )
prompts: govern only config files under mcp/ directories (87f3fcc )
quality: count && and || in Go cyclomatic complexity (30b7be1 )
quality: keep legacy Go cyclomatic counting as the 1.x default (8e59a2a )
restore 1.x backward compatibility for MCP, SDK risk weights and Go complexity (#195 ) (59f8e0c )
rules: retire dead quality.mixed-abstraction-levels; require emission sites (M6) (bd3c736 )
runner: fail scans whose target path is missing or unreadable (00a6a48 )
runner: fail the scan when a check section panics (35349a5 )
runner: report identical findings from overlapping targets once (76405d6 )
runner: stop callers mutating the shared corpus listing; make suppressed order total (062b01d )
sdk: restore int risk weights and track explicit 0 per key (bcdf068 )
security: close Python taint gaps for request data, imported sinks (5593e90 )
security: grade provider-prefixed keys at high confidence (2efe169 )
security: match insecure-deserialization on masked source (ae38101 )
security: stop scanArgLiterals panicking on a lone-quote argument (a6b1615 )
security: stop the secret name heuristic spanning concatenation (26b4212 )
stop codeguard silently passing on broken scans (high-priority review fixes) (#189 ) (4104604 )
Performance Improvements
diff: batch and memoize base-ref reads in diff mode (04a7efb )
quality: compute TS/JS module-scope flags in a single pass (310df11 )
quality: cut self-scan time ~46% by fixing three profiled hot paths (#191 ) (7660755 )
quality: gate the style-profile regexes behind literal pre-checks (4852024 )
quality: replace the clone tokenizer regex with a byte lexer (1e24245 )
quality: run whole-target passes concurrently and parallelize their per-file loops (9170132 )
runner: memoize C-like parses and inline suppressions per scan (e3095fe )
scan: derive scan concurrency from GOMAXPROCS and parallelise serial passes (b68b7e4 )
You can’t perform that action at this time.