Riqwan Thamir
|
ff975f1b9e
|
perf(core): Stop writing Instance AI skills to the sandbox (no-changelog) (#41220)
CI: Master (Build, Test, Lint) / Build & format (push) Waiting to run
CI: Master (Build, Test, Lint) / Unit tests (24.18.1) (push) Waiting to run
CI: Master (Build, Test, Lint) / Unit tests (26.5.1) (push) Waiting to run
CI: Master (Build, Test, Lint) / Lint (push) Waiting to run
CI: Master (Build, Test, Lint) / Verify single-instance deps (npm-install) (push) Waiting to run
CI: Master (Build, Test, Lint) / Notify Slack on failure (push) Blocked by required conditions
Co-authored-by: Cursor <cursoragent@cursor.com>
|
2026-10-17 14:02:07 +00:00 |
|
Raúl Gómez Morales
|
3c7dcf902d
|
feat(ai-builder): Write LangTracer results from the routing evals (no-changelog) (#40954)
|
2026-10-14 15:42:42 +00:00 |
|
José Braulio González Valido
|
1b0cdd5998
|
fix(ai-builder): Report eval helper model failures as infrastructure, not as bad JSON (no-changelog) (#40415)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-10-14 11:23:25 +00:00 |
|
Jaakko Husso
|
726a5829c8
|
fix(core): Use the module registry to decide if the Assistant has the Agents features (#41026)
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
|
2026-10-14 08:19:48 +00:00 |
|
Bernhard Wittmann
|
ccf42e3213
|
fix(ai-builder): Make browser credential-setup evals grade the right thing (no-changelog) (#40858)
|
2026-10-14 06:24:28 +00:00 |
|
José Braulio González Valido
|
29de1c63ae
|
fix(ai-builder): Remove the Context7 docs lookup from the eval mock (no-changelog) (#40021)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-10-13 21:30:08 +00:00 |
|
José Braulio González Valido
|
146d7804af
|
fix(ai-builder): Count eval runs that read another build's workflow as framework issues (no-changelog) (#40717)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-10-13 21:29:00 +00:00 |
|
José Braulio González Valido
|
88de3cac20
|
feat(ai-builder): Move eval judges to Sonnet 5.5 and add per-role eval LLM settings (no-changelog) (#40330)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-10-13 21:28:25 +00:00 |
|
José Braulio González Valido
|
5480258fb5
|
feat(ai-builder): Report the eval harness's own model usage per case (no-changelog) (#40718)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-10-09 07:52:09 +00:00 |
|
José Braulio González Valido
|
5b78e8b655
|
fix(ai-builder): Run each eval build in its own empty project (no-changelog) (#40715)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-10-08 21:15:11 +00:00 |
|
Raúl Gómez Morales
|
e67e5a9ef1
|
test(core): Run routing cases that start with state (no-changelog) (#40378)
|
2026-10-08 15:33:58 +00:00 |
|
Raúl Gómez Morales
|
6b8ac76d96
|
test(core): Add a routing mode to the Assistant discovery evals (no-changelog) (#40288)
|
2026-10-08 07:55:26 +00:00 |
|
José Braulio González Valido
|
0e1c754999
|
fix(ai-builder): Take feature flags from env only on Instance AI eval instances (no-changelog) (#40117)
CI: Master (Build, Test, Lint) / Build & format (push) Waiting to run
CI: Master (Build, Test, Lint) / Unit tests (24.18.1) (push) Waiting to run
CI: Master (Build, Test, Lint) / Unit tests (26.5.1) (push) Waiting to run
CI: Master (Build, Test, Lint) / Lint (push) Waiting to run
CI: Master (Build, Test, Lint) / Verify single-instance deps (npm-install) (push) Waiting to run
CI: Master (Build, Test, Lint) / Notify Slack on failure (push) Blocked by required conditions
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-10-06 20:46:09 +00:00 |
|
Robin Braumann
|
ea9021ed86
|
feat(core): Add durable pause checkpoints for background sub-agents (no-changelog) (#40118)
|
2026-10-02 13:10:50 +00:00 |
|
Anne Aguirre
|
3745c835b2
|
feat(core): Add a guardrail runner to the agent loop (#39621)
Co-authored-by: Cursor <cursoragent@cursor.com>
|
2026-10-01 10:21:42 +00:00 |
|
José Braulio González Valido
|
1b1bb4698d
|
fix(ai-builder): Bind a seeded eval Agent's model to the case credential (no-changelog) (#39984)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-09-30 22:18:59 +00:00 |
|
José Braulio González Valido
|
e185c36c59
|
fix(ai-builder): Clear eval dedupe state between runs and add a Sheets mock quirk (no-changelog) (#39990)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-09-30 22:18:56 +00:00 |
|
José Braulio González Valido
|
0c8acfa274
|
fix(ai-builder): Send requiresMemoryCompaction on eval case push (no-changelog) (#39989)
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
|
2026-09-30 17:34:50 +00:00 |
|
José Braulio González Valido
|
54f634d5f9
|
fix(ai-builder): Cancel the thread when an eval turn budget fires after the run (no-changelog) (#39992)
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
|
2026-09-30 17:34:47 +00:00 |
|
Elias Meire
|
202a52d206
|
fix: Update broken external links in core packages (no-changelog) (#39704)
|
2026-09-30 12:41:50 +00:00 |
|
yehorkardash
|
99182c29be
|
feat(core): Change IAI permissions logic (#39303)
|
2026-09-30 09:32:20 +00:00 |
|
José Braulio González Valido
|
995c101dba
|
fix(ai-builder): Read the reseeded scenario tables live instead of from pinned rows (no-changelog) (#39814)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-09-29 13:53:20 +00:00 |
|
José Braulio González Valido
|
bf6ba26a7f
|
feat(ai-builder): Budget eval builds per user turn and make a timeout its own outcome (no-changelog) (#39540)
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
|
2026-09-29 12:52:52 +00:00 |
|
José Braulio González Valido
|
918512f62a
|
fix(ai-builder): Keep scenario tables for Agent builds and reuse same-named seed tables (no-changelog) (#39767)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
|
2026-09-28 15:40:32 +00:00 |
|
Riqwan Thamir
|
c1d666301f
|
feat(editor): Make Instance AI LLM step debug data readable (no-changelog) (#39575)
Co-authored-by: Cursor <cursoragent@cursor.com>
|
2026-09-28 15:11:01 +00:00 |
|
Michael Drury
|
8a0ca48bb7
|
feat(core): Add shared Agent context access (#38930)
|
2026-09-25 14:26:57 +00:00 |
|
José Braulio González Valido
|
e3fe95d228
|
fix(ai-builder): Honor neverError in eval mocks, seed agent tables, filter pinned reads and report start refusals (no-changelog) (#39337)
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
|
2026-09-25 07:47:28 +00:00 |
|
José Braulio González Valido
|
aac074ab06
|
fix(core): Let the eval user proxy complete a templated-auth credential card (no-changelog) (#39115)
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
|
2026-09-25 00:19:15 +00:00 |
|
José Braulio González Valido
|
44471d1c94
|
fix(ai-builder): Validate only the selected case files in the eval push CLI (no-changelog) (#39248)
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
|
2026-09-24 20:11:04 +00:00 |
|
oleg
|
0254343de3
|
feat(core): Track n8n Assistant workflow setup (no-changelog) (#39190)
|
2026-09-23 10:37:44 +00:00 |
|
Arvin A
|
724f1f7f90
|
feat(core): Show the injected instance context in the trace (no-changelog) (#38240)
|
2026-09-22 17:15:29 +00:00 |
|
José Braulio González Valido
|
5b149495c5
|
feat(ai-builder): Grade token consumption and memory compaction in Instance AI evals (no-changelog) (#38974)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
2026-09-22 11:12:10 +00:00 |
|
José Braulio González Valido
|
ecda0d4680
|
fix(ai-builder): Report un-intercepted vendor sub-nodes and stop blaming the mock layer for a root's own crash (no-changelog) (#38155)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
2026-09-21 12:18:30 +00:00 |
|
Arvin A
|
d9dc457d40
|
feat(core): Track credential description usage (no-changelog) (#38986)
|
2026-09-21 09:09:43 +00:00 |
|
Michael Drury
|
d605e8ca80
|
test(core): Support Agent attachments in Instance AI evaluations (#38874)
|
2026-09-21 08:57:39 +00:00 |
|
José Braulio González Valido
|
326030418d
|
fix(core): Attribute a draft-Agent eval scenario to the harness, not the builder (no-changelog) (#39007)
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
|
2026-09-18 11:44:52 +00:00 |
|
Riqwan Thamir
|
0a4976e62c
|
feat(editor): Open Edit with AI without a greeting turn (#38745)
Co-authored-by: Cursor <cursoragent@cursor.com>
|
2026-09-17 16:45:07 +00:00 |
|
José Braulio González Valido
|
e15b3fc922
|
fix(core): Restore the system prompt in Instance AI LLM step debug (no-changelog) (#38887)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
2026-09-17 13:26:23 +00:00 |
|
Jaakko Husso
|
ac752e7026
|
feat(core): Add an AI Assistant tool that runs one node of a workflow (#38601)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
2026-09-16 14:16:44 +00:00 |
|
Roman Makowiak
|
5b199f5cdf
|
feat(core): Add folder seeding to the Instance AI eval harness (no-changelog) (#38550)
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
|
2026-09-16 07:47:46 +00:00 |
|
Mutasem Aldmour
|
2f352bde2b
|
fix(core): Offer Computer Use only where the client shows its entry points (#38108)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
2026-09-15 15:35:01 +00:00 |
|
Jaakko Husso
|
7b10a16ca4
|
feat(core): Tell Instance AI when a verified draft is not the live workflow (#38289)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
2026-09-14 13:47:19 +00:00 |
|
Michael Drury
|
684b930fa6
|
feat(core): Improve Agent Builder evaluation and intent routing (#38329)
|
2026-09-14 13:13:51 +00:00 |
|
Luka Zivkovic
|
70e5a2bf16
|
feat: Export structured agent artifacts in eval results (#38394)
|
2026-09-11 15:11:02 +00:00 |
|
Raúl Gómez Morales
|
e44fbd358a
|
fix(core): Return Instance AI node output grouped per output branch (#38414)
|
2026-09-11 11:51:35 +00:00 |
|
oleg
|
1bee3bcac7
|
feat(core): Add progressive workflow building (no-changelog) (#37996)
|
2026-09-10 13:42:43 +00:00 |
|
Matsu
|
7be94a1184
|
chore: Update eslint configuration (#38307)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
2026-09-10 14:40:35 +03:00 |
|
José Braulio González Valido
|
44fae2aac9
|
fix(ai-builder): Give agent narration a larger judge budget than tool payloads (no-changelog) (#38192)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
2026-09-10 08:37:07 +00:00 |
|
Raúl Gómez Morales
|
564646cfac
|
feat(core): Add published flag to Instance AI eval seed workflows (no-changelog) (#38209)
|
2026-09-10 08:15:02 +00:00 |
|
oleg
|
dd0e1f61bc
|
feat(core): Add setup panel verification and agent test execution (no-changelog) (#38031)
|
2026-09-10 07:46:46 +00:00 |
|