v0.8.0

0.8.0 (2026-10-10)

Features

• brief: propose audience-grounded angles during research (#142) (7ade888)
• discovery: rank suggested topics by audience fit (#143) (3ac904f)
• governance: suggest setup changes from recurring review corrections (#144) (f0850d3)
• tenant: control company mentions and carry brief direction into review (#139) (35aeb97)

Bug Fixes

• tenant: address review findings across brief angles, discovery, and suggestions (#145) (226b4e3)
• tenant: separate "Not our user" and churn triggers (#140) (5905ae6)

Performance Improvements

• cut admin and pipeline query load; fix audit and job-queue bugs (#122) (211267c)

v0.7.0

0.7.0 (2026-09-14)

Features

• cms: polish the operator journey, consolidate surfaces, remove legacy code (#110) (7bb204f)

v0.6.4

0.6.4 (2026-09-11)

Bug Fixes

• Improve ops workspace guidance and review layouts (#98) (8e97da5)

v0.6.3

0.6.3 (2026-09-05)

Bug Fixes

• cms: aggregate report costs in Postgres (#93) (27d6923)
• cms: queue concurrent selected content runs reliably (#89) (aba9d4f)
• cms: refresh articles when short runs finish between polls (#90) (b68a0fd)
• cms: restore GraphQL with valid policy enum options (#96) (d57f27c)
• cms: support keyboard selection of content templates (#91) (bda8404)
• cms: validate signed cache invalidation requests (#95) (94b0a8c)
• pipeline: report progress only after article writes succeed (#87) (f07b44c)
• pipeline: report saved results without research setup (#88) (bf30442)

v0.6.2

0.6.2 (2026-09-04)

Performance Improvements

• cms: speed up content lists and reduce admin data loading (#85) (6e9cc1c)

v0.6.1

0.6.1 (2026-09-04)

Bug Fixes

• cms: Validate models before content discovery (715bbf8)
• security: Disable live local Codex execution (f15f5a2)
• security: Disable live local Codex execution (f9fb74b)

v0.6.0

0.6.0 (2026-09-03)

Features

• onboarding: add tenant context assets and AI-assisted setup (42b7c20)
• onboarding: add tenant context assets and AI-assisted setup (e3275f3)

Bug Fixes

• evidence: harden evidence bank, site crawl, and demo fill after review (e489593)

v0.5.0

0.5.0 (2026-09-03)

Features

• cms: add a Game Preview template (7d25dac)
• llm: run pipeline stages on a ChatGPT plan via the Codex CLI (#67) (62090b8)
• pipeline: carry the template example into the generate prompt (54fef48)
• templates: add Game Preview and carry the example into the generate prompt (72ca7b4)

Bug Fixes

• templates: order the Game Preview example like its outline (faa96a9)

v0.4.0

0.4.0 (2026-08-31)

Features

• ops: webhooks, scheduled publishing, caching, and a shared status table (#63) (98db434)

v0.3.0

0.3.0 (2026-08-27)

Features

• cms: add the evidence-source-candidates collection (35f7fc8)
• cms: add the source review governance page (015aff4)
• cms: add the topic discovery migration (b289188)
• cms: group picked topics into one article and cache discovery (cdf1d5f)
• cms: intent-first new content flow that starts research on its own (5df5866)
• cms: let operators discover topics from a subject they type (1e12678)
• cms: onboarding is one decision, then your first piece (278a0fa)
• cms: replace the article board with a content list and stepper (9ad44e3)
• cms: report run outcomes on the board and allow new templates (a35393d)
• cms: run and remove selected topics from the article board (cf03720)
• cms: say what QA failed and what the rewrite must change (1ec050b)
• cms: say what the model is doing during a live run (3a2b758)
• cms: show active runs in a sticky bar across the admin (57cacf2)
• pipeline: add the brief checkpoint between research and writing (bda52a7)
• pipeline: record unrated evidence domains from the scoring stage (c32951b)

Bug Fixes

• cms: keep the domain rating pill on one line (b32e575)
• cms: make the source review page match the ops CSS conventions (c9f35b6)
• cms: Prevent integration schema races (9dd2e14)
• cms: unstick deferred research and fix two Codex findings (5125ef1)
• pipeline: label FAQ entries so QA stops reporting them as duplicate content (e7b6417)
• pipeline: repair the OpenAI path and the checks it feeds (013b19f)
• pipeline: stop failing every draft on a missing social image (116aee8)
• pipeline: stop the e2e report check failing when it finds the block (f20cb34)
• pipeline: stop truncating command output when it is piped (fc7c26c)
• pipeline: update workspaceReadiness tests for the ready redefinition (b157e20)