Commit Graph
410 Commits
Author SHA1 Message Date
Dotta 49e06605d8 Add serialized two-lane chat campaign coordinator 2026-10-27 04:42:56 -05:00
Dotta d476ef75bf Retain first ten efbb initial measurements and live checkpoint 2026-10-27 04:39:17 -05:00
Dotta 0be6d627d0 test(evals): complete protected human GitHub grant continuation 2026-10-27 04:36:48 -05:00
Dotta 5c9eaa8ee6 test(e2e): accept explicit broad ChatGPT write consent 2026-10-27 04:24:44 -05:00
Dotta b349eb24c4 test(e2e): author genuine named ChatGPT human setup 2026-10-27 04:23:01 -05:00
Dotta ccf69b1743 Reconcile frozen catalog and retain first 43 measurements 2026-10-27 04:20:24 -05:00
Dotta f457fa84b0 Author private Grok operator OAuth and pairing handoff 2026-10-27 04:11:42 -05:00
Dotta 406eb610dd test(evals): declare exact private ChatGPT human setup prerequisites 2026-10-27 04:02:04 -05:00
Dotta 61a8c7e4ae test(evals): close genuine team installer conflict and partial outcomes 2026-10-27 03:59:17 -05:00
Dotta 464675d888 Declare explicit selected Grok human operator prerequisites 2026-10-27 03:57:42 -05:00
Dotta 8183b6d7a3 test(evals): author natural Notion expiry cancellation and real reconnect 2026-10-27 03:43:30 -05:00
Dotta 21a19b06f3 fix(evals): bind instruction edits to supported file mutation audit 2026-10-27 03:41:37 -05:00
Dotta 56eb41b819 fix(evals): keep repeated run observations out of browser trace 2026-10-27 03:41:37 -05:00
Dotta aae4f384e0 fix(evals): resolve only anchored complete cleanup evidence 2026-10-27 03:36:27 -05:00
Dotta c5d20de8f5 test(evals): prove personal Notion workspace selection preserves shared access 2026-10-27 03:30:55 -05:00
Dotta 77d64e17cc test(evals): author real Notion OAuth handoff and draft recovery 2026-10-27 03:26:22 -05:00
Dotta 2e287f0f61 Author real two-message review reply and verified-fix phases 2026-10-27 03:24:58 -05:00
Dotta 580c17fd87 fix(evals): complete factual setup context before held queries 2026-10-27 03:24:22 -05:00
Dotta bf97b63757 docs(evals): retain continued initial measurements and held corrections 2026-10-27 03:21:52 -05:00
Dotta 200e7fc3ac fix(evals): bind held UI journeys to actual controls 2026-10-27 03:20:56 -05:00
Dotta 983f15394e fix(evals): preserve catalog intent in held correction set 2026-10-27 03:18:47 -05:00
Dotta 0f13d2643a Author actual launch email and authentic receipt duplicate journey 2026-10-27 03:07:33 -05:00
Dotta 62e31c5632 Author real Slack authorization expiry and normal human reconnect proof 2026-10-27 03:05:19 -05:00
Dotta 81d12acc6e Checkpoint first passed initial and unaffected sequential batch 2026-10-27 02:52:20 -05:00
Dotta 61c863364a Require distinct saved tasks for broken-link fixture 2026-10-27 02:51:10 -05:00
Dotta 067f1ecddf Bind uncertain-email setup to genuine normal-task native authority 2026-10-27 02:49:55 -05:00
Dotta b26d4f7df0 Retain broken-link fixture and diagnostic privacy failure 2026-10-27 02:46:44 -05:00
Dotta fd356aa4f8 Select authorized AgentMail targets and private sender refs for email journeys 2026-10-27 02:44:26 -05:00
Dotta b21eeaa239 Author normal Slack setup with private human slash command and real wizard activation 2026-10-27 02:39:51 -05:00
Dotta eeccfd9bde Install exact selected uncertain-email transport fault before service creation 2026-10-27 02:36:40 -05:00
Dotta cbb6593441 Download registered artifacts through authenticated attachment transport 2026-10-27 02:36:26 -05:00
Dotta 12130d11b6 Author genuine uncertain-email recovery and scoped response loss 2026-10-27 02:36:05 -05:00
Dotta 1fc30a0ed1 Validate exact owned database config custody in persisted-state scans 2026-10-27 02:24:09 -05:00
Dotta f2a4123b8d Retain comment startup failure and stop remaining admissions 2026-10-27 02:17:59 -05:00
Dotta 8ebabf9a8e Add real Slack sender confirmation and current resource boundary eval 2026-10-27 02:16:24 -05:00
Dotta 96058c7e2f Keep owned Docker database credentials in server-only environment 2026-10-27 02:11:39 -05:00
Dotta 1454f54399 Record project initial privacy failure and halt admissions 2026-10-27 02:07:51 -05:00
Dotta f6b2d50c84 Author owned lost-sandbox reconstruction chat evaluation 2026-10-27 01:58:20 -05:00
Dotta 82228ed909 Use complete public run identity in note oracle controls 2026-10-27 01:53:16 -05:00
Dotta 55568965be test(evals): require positive vault import despite provider diagnostics 2026-10-27 01:52:17 -05:00
Dotta 46de5b733e Reconcile 229 authored catalog paths from immutable source checkpoint 2026-10-27 01:50:48 -05:00
Dotta 7fb1b10dea Bind note negation to the referenced design instruction 2026-10-27 01:50:48 -05:00
Dotta 2545a5d534 Require affirmative saved note bound to the actual design link 2026-10-27 01:48:28 -05:00
Dotta 15ce55696e test(evals): author human vault import and native reference use 2026-10-27 01:47:56 -05:00
Dotta 5280ce12fb Author genuine retained workspace repair chat evaluation 2026-10-27 01:45:57 -05:00
Dotta 2e6a5dbe27 Keep broken-link prerequisite descriptions factual 2026-10-27 01:43:45 -05:00
Dotta 32b472dfa5 Refresh reviewed MCP SDK dynamic source binding 2026-10-27 01:42:10 -05:00
Dotta 8f98db3d21 test(evals): author task repository preparation and explicit grant gap 2026-10-27 01:40:36 -05:00
Dotta 4afcd710e2 Bind selected public sign-in worker and actual auth UI source 2026-10-27 01:36:24 -05:00
Dotta ea6dc42d2b Add finite owned public HTTPS origin handoff for chat evals 2026-10-27 01:31:43 -05:00