Commit Graph

  • 5294c65777 fix(eval): handle SDK max-turns throw gracefully Garry Tan 2026-04-23 11:34:51 -07:00
  • 416a56a5c8 fix(ship): exclude current PR from queue-awareness (self-reference bug) Garry Tan 2026-04-23 11:28:02 -07:00
  • bd39b6995f chore: bump version and changelog (v1.8.0.0) Garry Tan 2026-04-23 11:12:42 -07:00
  • 6efd6483e8 Merge remote-tracking branch 'origin/main' into garrytan/workspace-aware-ship Garry Tan 2026-04-23 11:08:17 -07:00
  • cb722a7fa7 docs: versioning invariant in CLAUDE.md Garry Tan 2026-04-23 11:08:12 -07:00
  • d7fa332803 feat(skill): /landing-report read-only queue dashboard Garry Tan 2026-04-23 11:08:12 -07:00
  • f4ec92341c feat(skills): queue-aware /ship + drift abort in /land-and-deploy + advisory in /review Garry Tan 2026-04-23 11:08:12 -07:00
  • 04e2f1bea9 test(eval): 3 more overlay fixtures to measure remaining Claude nudges Garry Tan 2026-04-23 11:08:09 -07:00
  • c46388ac4d feat(ci): version-gate + pr-title-sync workflows (GitHub + GitLab) Garry Tan 2026-04-23 11:07:57 -07:00
  • c7bdfdf304 feat(scripts): detect-bump + compare-pr-version helpers Garry Tan 2026-04-23 11:07:56 -07:00
  • 236e9d91cc test: fixture tests for gstack-next-version Garry Tan 2026-04-23 11:07:56 -07:00
  • 7ed110e57b feat: bin/gstack-next-version util + workspace_root config key Garry Tan 2026-04-23 11:07:46 -07:00
  • 76829b76dc feat(eval): extend OverlayFixture with allowedTools, maxTurns, direction Garry Tan 2026-04-23 11:07:37 -07:00
  • f3b40b12d7 Merge remote-tracking branch 'origin/main' into garrytan/overlay-fanout-eval Garry Tan 2026-04-23 10:54:42 -07:00
  • 8a3f197aea chore: bump version and changelog (v1.6.5.0) Garry Tan 2026-04-23 10:37:25 -07:00
  • 54d3ad923d Merge branch 'main' into garrytan/gbrain-support Garry Tan 2026-04-23 10:36:30 -07:00
  • d75402bbd2 v1.6.4.0: cut Haiku classifier FP from 44% to 23%, gate now enforced (#1135) Garry Tan 2026-04-23 10:23:40 -07:00
  • bfb8ac3b7f Merge remote-tracking branch 'origin/main' into garrytan/overlay-fanout-eval Garry Tan 2026-04-23 09:14:16 -07:00
  • cb5f074d7c fix(opus-4.7): remove "Fan out explicitly" overlay nudge Garry Tan 2026-04-23 09:14:11 -07:00
  • 546404c81f test: register overlay harness in touchfiles (both maps) Garry Tan 2026-04-23 09:13:59 -07:00
  • e432f4bd94 test(eval): paid periodic overlay-efficacy harness Garry Tan 2026-04-23 09:13:58 -07:00
  • 06a862faab test(eval): unit tests for agent-sdk-runner (36 tests, free tier) Garry Tan 2026-04-23 09:13:58 -07:00
  • 6b85262422 feat(eval): parametric overlay-efficacy harness (runner + fixtures) Garry Tan 2026-04-23 09:13:58 -07:00
  • 66887b2f05 feat(preflight): sanity check for agent-sdk + overlay resolver Garry Tan 2026-04-23 09:13:37 -07:00
  • 27e4ee7498 chore: add @anthropic-ai/claude-agent-sdk@0.2.117 dep Garry Tan 2026-04-23 09:13:37 -07:00
  • 6c6fa69191 refactor: export readOverlay from model-overlay resolver Garry Tan 2026-04-23 09:13:37 -07:00
  • c7d6add473 fix(test): session-awareness reads AskUserQuestion Format from a Tier 2+ SKILL.md Garry Tan 2026-04-23 09:10:59 -07:00
  • f43369a453 docs(changelog): rewrite v1.6.4.0; strip process minutiae #1135 garrytan/injection-tuning Garry Tan 2026-04-23 09:09:42 -07:00
  • aa6d69a0a8 docs(changelog): strip process minutiae from entries; rewrite v1.6.4.0 Garry Tan 2026-04-23 08:29:38 -07:00
  • a882c05151 docs(changelog): add v1.6.4.0 placeholder entry at top Garry Tan 2026-04-23 07:34:09 -07:00
  • 8407b4930b Merge branch 'main' into garrytan/gbrain-support Garry Tan 2026-04-23 07:30:58 -07:00
  • f94ee63460 merge: origin/main into garrytan/injection-tuning; bump v1.6.2.0 → v1.6.4.0 Garry Tan 2026-04-23 07:28:41 -07:00
  • 69733e2622 fix(plan-reviews): restore RECOMMENDATION + Completeness split + Codex ELI10 (v1.6.3.0) (#1149) Garry Tan 2026-04-23 07:25:20 -07:00
  • 62fa71962c chore: bump version and changelog (v1.6.3.0) #1149 garrytan/plan-review-regressions Garry Tan 2026-04-23 07:09:13 -07:00
  • 09c82222ea test: fix Codex eval sandbox + collector API Garry Tan 2026-04-22 22:02:32 -07:00
  • 028627fbcd fix(preamble): harden AskUserQuestion Format + Codex ELI10 carve-out Garry Tan 2026-04-22 21:34:31 -07:00
  • b7f6246061 test: add Codex eval for AskUserQuestion format compliance Garry Tan 2026-04-22 21:34:11 -07:00
  • 12b29c7518 merge: origin/main into garrytan/injection-tuning; bump v1.5.2.0 → v1.6.2.0 Garry Tan 2026-04-22 17:00:33 -07:00
  • 756525100c chore: regenerate SKILL.md files for gbrain-sync preamble block Garry Tan 2026-04-22 13:58:41 -07:00
  • c3f73f91d4 chore: bump version and changelog (v1.7.0.0) Garry Tan 2026-04-22 13:47:12 -07:00
  • 91c734a6af docs(gbrain-sync): user guide + error lookup + README section Garry Tan 2026-04-22 13:47:12 -07:00
  • c064743eda test(gbrain-sync): 27-test consolidated suite Garry Tan 2026-04-22 13:47:12 -07:00
  • a2aa8a07d4 feat(gbrain-sync): preamble block — privacy gate + boundary sync Garry Tan 2026-04-22 13:47:12 -07:00
  • f088fe96f8 feat(gbrain-sync): init, restore, uninstall, consumer registry Garry Tan 2026-04-22 13:47:11 -07:00
  • 97cbacf409 feat(gbrain-sync): --once drain + secret scan + push Garry Tan 2026-04-22 13:47:11 -07:00
  • 45638297ba feat(gbrain-sync): queue primitives + writer shims Garry Tan 2026-04-22 13:47:11 -07:00
  • d591ad29b2 chore: bump version and changelog (v1.6.2.0) Garry Tan 2026-04-22 12:32:14 -07:00
  • 00e8a8599c Merge remote-tracking branch 'origin/main' into garrytan/plan-review-regressions Garry Tan 2026-04-22 12:29:35 -07:00
  • 5fe1814310 fix(plan-reviews): restore RECOMMENDATION + split Completeness by question type Garry Tan 2026-04-22 01:11:05 -07:00
  • 6a46b9099f test: add AskUserQuestion format regression eval for plan reviews Garry Tan 2026-04-22 01:10:35 -07:00
  • 656df0e37e feat(v1.5.2.0): Opus 4.7 migration — model overlay, voice, routing (#1117) Garry Tan 2026-04-22 01:06:22 -07:00
  • 03322352dd test(opus-4.7): key touchfile entries by testName, not describe text #1117 feat/opus-4.7-migration Garry Tan 2026-04-22 00:37:27 -07:00
  • 206bf93390 chore(release): v1.6.1.0 — Opus 4.7 migration, reviewed Garry Tan 2026-04-22 00:30:38 -07:00
  • 404c9c925d docs(todos): verify Opus 4.7 fanout nudge in Claude Code harness (P0) Garry Tan 2026-04-22 00:28:54 -07:00
  • d646bc12d8 test(opus-4.7): rewrite scratch-root helper + add afterAll cleanup Garry Tan 2026-04-22 00:27:53 -07:00
  • 723f9957f2 test(opus-4.7): tighten ambiguous /qa routing prompt Garry Tan 2026-04-22 00:27:28 -07:00
  • 36ef9d9db0 refactor(opus-4.7): rewrite fanout nudge to show parallel tool_use pattern Garry Tan 2026-04-22 00:26:26 -07:00
  • 7e90b0f092 test(opus-4.7): E2E eval for fanout rate + routing precision Garry Tan 2026-04-22 00:11:38 -07:00
  • d3742c884a test(team-mode): give setup -q / setup --local tests a 3-minute budget Garry Tan 2026-04-21 23:48:48 -07:00
  • b6be59ab75 test(binary-guard): replace xargs-per-file loops with fs.statSync + mode filter Garry Tan 2026-04-21 23:43:43 -07:00
  • b90bc40295 test(routing): assert slash-prefixed skills + new policy + current names Garry Tan 2026-04-21 23:41:31 -07:00
  • 6701205118 chore(opus-4.7): regenerate SKILL.md files + update golden fixtures Garry Tan 2026-04-21 23:40:07 -07:00
  • da75ebaaa0 refactor(opus-4.7): split overlay, align routing, fix trailer fallback Garry Tan 2026-04-21 23:39:42 -07:00
  • 4d260db78a feat(v1.5.2.0): Opus 4.7 migration — model overlay, voice, routing gstack 2026-04-21 05:57:02 +00:00
  • 54d4cde773 security: tunnel dual-listener + SSRF + envelope + path wave (v1.6.0.0) (#1137) Garry Tan 2026-04-21 21:58:27 -07:00
  • 89480291f5 fix(sidebar): killAgent resets per-tab state; align tests with current agent event format #1137 garrytan/sec-wave Garry Tan 2026-04-21 21:34:30 -07:00
  • 499f2ade92 fix(make-pdf): write --from-file payload to /tmp, not os.tmpdir() Garry Tan 2026-04-21 21:19:44 -07:00
  • 3bff673671 docs: update project documentation for v1.6.0.0 Garry Tan 2026-04-21 21:11:55 -07:00
  • c7583f7f02 fix: pre-landing review findings (4 auto-fixes) Garry Tan 2026-04-21 21:07:35 -07:00
  • 596a130d60 merge: origin/main into garrytan/injection-tuning; bump v1.5.1.0 → v1.5.2.0 Garry Tan 2026-04-21 20:50:02 -07:00
  • cb71b72ce3 release(v1.6.0.0): bump VERSION + CHANGELOG for security wave Garry Tan 2026-04-21 20:48:52 -07:00
  • f5ad22c3bb test(E1): end-to-end pair-agent flow against a spawned daemon Garry Tan 2026-04-21 20:45:31 -07:00
  • 580f54d3e1 docs(E2): document dual-listener tunnel architecture in ARCHITECTURE.md Garry Tan 2026-04-21 20:43:18 -07:00
  • 80dfc988a1 security(N2): document Windows v20 ABE elevation path on CDP port Garry Tan 2026-04-21 20:42:15 -07:00
  • 13a7528697 security(N1): replace ?token= SSE auth with HttpOnly session cookie Garry Tan 2026-04-21 20:41:07 -07:00
  • 49263b3d10 security(E3): gate GSTACK_SLUG on /welcome path traversal Garry Tan 2026-04-21 20:37:37 -07:00
  • 202b4308b1 fix(scripts): drop shell:true from slop-diff npx invocations Mohammed Qazi 2026-04-18 15:59:56 -07:00
  • de4832ba1a fix(design): escape url.origin when injecting into served HTML Mohammed Qazi 2026-04-18 15:59:49 -07:00
  • 693eadf6f3 security: validate --from-file payload paths for parity with direct paths gus 2026-04-20 14:28:29 -03:00
  • 1372a4f631 security: extend hidden-element detection to all DOM-reading channels gus 2026-04-16 20:33:09 -03:00
  • d9e78dd548 security: route splitForScoped through envelope sentinel escape gus 2026-04-16 19:58:31 -03:00
  • b7cf46d6e6 security: gate download + scrape through validateNavigationUrl (SSRF) gus 2026-04-16 19:47:19 -03:00
  • 7dd5dbba83 chore(release): v1.5.1.0 — cut Haiku FP 44% → 23% Garry Tan 2026-04-21 20:32:24 -07:00
  • b86ef06706 test(security): live + fixture-replay bench harness with 500-case capture Garry Tan 2026-04-21 20:32:15 -07:00
  • de972b514f test(security): add source-level guards for dual-listener architecture Garry Tan 2026-04-21 20:31:56 -07:00
  • 6cedecd585 feat(security): v2 ensemble tuning — label-first voting + SOLO_CONTENT_BLOCK Garry Tan 2026-04-21 20:31:53 -07:00
  • 67a511c0b1 feat(security): dual-listener tunnel architecture Garry Tan 2026-04-21 20:31:43 -07:00
  • f962796f07 feat(security): add tunnel-denial-log module for attack visibility Garry Tan 2026-04-21 20:31:14 -07:00
  • 12fdc6391c refactor(security): loosen /connect rate limit from 3/min to 300/min Garry Tan 2026-04-21 20:31:03 -07:00
  • dcc820f070 fix(stealth): address adversarial code review — 7 findings #1112 feature/stealth-patches gstack 2026-04-21 03:10:13 +00:00
  • 1eae837260 fix(stealth): address security review findings gstack 2026-04-21 03:08:06 +00:00
  • 8df1c003b5 test(stealth): comprehensive unit + e2e tests, fix platform mismatch gstack 2026-04-21 03:06:40 +00:00
  • 46b2e35917 feat(browse): comprehensive anti-bot stealth patches gstack 2026-04-21 02:45:00 +00:00
  • e23ff280a1 fix(v1.4.1.0): /make-pdf — page numbers, entity escape, Linux fonts (#1098) Garry Tan 2026-04-20 22:32:58 +08:00
  • 109994b6b0 Merge remote-tracking branch 'origin/main' into garrytan/makepdf-patch #1098 garrytan/makepdf-patch Garry Tan 2026-04-20 22:22:21 +08:00
  • 72d94c3df3 chore: anonymize test fixtures, drop VC-partner framing Garry Tan 2026-04-20 22:19:55 +08:00
  • 97584f9a59 feat(security): ML prompt injection defense for sidebar (v1.4.0.0) (#1089) Garry Tan 2026-04-20 22:18:37 +08:00
  • 2b4a2c7840 test(security): assert block stops further tool calls, allow lets them through #1089 garrytan/prompt-injection-guard Garry Tan 2026-04-20 22:03:29 +08:00
  • 53202b5a14 docs(todos): P0 Haiku FP tuning + P1-P3 follow-ups from bench data Garry Tan 2026-04-20 22:00:16 +08:00
  • 372a5aa382 docs(changelog): measured BrowseSafe-Bench lift from Haiku unbreak Garry Tan 2026-04-20 21:57:16 +08:00