Skip to content

Releases: data-fair/agents

v0.9.2

Choose a tag to compare

@albanm albanm released this 29 Jun 18:33

What's Changed

Fixes 🔧

  • fix(traces): ship architecture docs in the production image by @albanm in #49
  • fix(subagents): recover a best-effort answer when a worker loops to its step cap by @albanm in #50

Full Changelog: v0.9.1...v0.9.2

v0.9.0

Choose a tag to compare

@albanm albanm released this 29 Jun 15:19

What's Changed

New Features 🎉

  • feat(subagents): stateless workers, concurrent panels, simplified display by @albanm in #48

Fixes 🔧

  • fix(scaleway): support project-scoped keys and chat-completions endpoint by @albanm in #45
  • fix(gateway): keep reasoning-model conversations from stalling by @albanm in #46
  • fix(traces): stop recording the evaluator's own review session by @albanm in #47

Full Changelog: v0.8.0...v0.9.0

v0.8.0

Choose a tag to compare

@albanm albanm released this 26 Jun 13:21

What's Changed

New Features 🎉

  • feat: add compatibility mode for openai-compatible providers by @albanm in #44

Full Changelog: v0.7.0...v0.8.0

v0.7.0

Choose a tag to compare

@albanm albanm released this 22 Jun 07:36

What's Changed

New Features 🎉

  • feat(traces): superadmin source-code context for the evaluator by @albanm in #35
  • feat(traces): superadmin evaluation via a promoted evaluator account by @albanm in #36
  • feat(traces): give the trace evaluator data-fair data-exploration tools by @albanm in #37
  • feat(moderation): give the input-moderation gate conversational context by @albanm in #39
  • feat(chat): render streamed markdown progressively, token by token by @albanm in #41
  • feat(chat): clearer streaming activity for agents and sub-agents by @albanm in #42

Fixes 🔧

  • fix(chat): stop conversations dropping silently or on over-eager moderation by @albanm in #38
  • fix(chat): prevent stalled turns from hanging and show activity status by @albanm in #40

Full Changelog: v0.6.0...v0.7.0

v0.6.0

Choose a tag to compare

@albanm albanm released this 19 Jun 12:20

What's Changed

New Features 🎉

  • feat(admin): nested account navigation with superadmin selector and scoped trace review by @albanm in #33
  • feat(traces): overhaul admin trace review (system-aware evaluator, costs, compare, scoped nav) by @albanm in #34

Full Changelog: v0.5.1...v0.6.0

v0.5.0

Choose a tag to compare

@albanm albanm released this 18 Jun 08:17

What's Changed

New Features 🎉

  • feat(chat-action): carry hidden context in the user turn, not the system prompt by @albanm in #24
  • feat(ui): render mermaid diagrams in chat (experimental, admin opt-in) by @albanm in #25
  • feat(chat): experimental flatten-subagents toggle with opt-out by @albanm in #26
  • feat(chat): open experimental flags to all users and tidy the settings tab by @albanm in #27
  • feat: bounded mermaid auto-fix and faithful tool-result handling by @albanm in #28
  • feat(traces): record experimental flags and add review summary bar by @albanm in #30

Fixes 🔧

  • fix(chat): resolve agent prose links against the host app base by @albanm in #23
  • fix(moderation): narrow the off-scope criterion to curb false positives by @albanm in #29
  • fix(chat): keep agent iframe link nav working when state created outside setup by @albanm in #31

Full Changelog: v0.4.0...v0.5.0

v0.4.0

Choose a tag to compare

@albanm albanm released this 11 Jun 12:49

What's Changed

New Features 🎉

  • feat(trace): capture physical LLM requests and richer sub-agent steps by @albanm in #7
  • feat(ui): denser and more consistent chat UI by @albanm in #8
  • feat(chat): cache metrics in debug traces, drop user name from prompt by @albanm in #9
  • feat(ui): add opt-in progressive tool disclosure via explore_tools by @albanm in #10
  • feat(gateway): require anonymous-action token on anonymous gateway/summary by @albanm in #12
  • feat(chat): autoscroll, sub-agent auto-open, jump-to-bottom button by @BatLeDev in #11
  • feat(ui): add settings tab with tool-exploration toggle to chat dialogyes by @albanm in #15
  • feat(limits): add untrusted pool quota for anonymous + external usage by @albanm in #16
  • feat(traces): opt-in consent-gated server-side conversation trace storage by @albanm in #18
  • feat(moderation): gateway-enforced untrusted moderation + observability by @albanm in #19
  • feat(moderation): per-org input moderation setting by @albanm in #22

Fixes 🔧

  • fix(gateway): preserve parallel tool calls in the streaming path by @albanm in #17
  • fix(summary): always pin the summarizer system prompt by @albanm in #21

Other Changes

  • docs: split architecture into topic files, add French presentation by @albanm in #20

New Contributors

Full Changelog: v0.3.0...v0.4.0

v0.3.0

Choose a tag to compare

@albanm albanm released this 29 May 12:44

What's Changed

New Features 🎉

  • feat(chat): support custom local chats in portal (long system prompt + flat render) by @albanm in #6

Full Changelog: v0.2.0...v0.3.0