mirror of
https://github.com/windmill-labs/windmill.git
synced 2026-10-05 08:02:28 +00:00
c5f59d6ccc71d08fc34ede47e7a77b35a197b9db
14950
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
  Ruben FiszelandClaude Opus 5.5
|
c5f59d6ccc |
fix: drop caller-supplied _MODULES at push for every job (#11418)
* fix: drop caller-supplied _MODULES at push for every job Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: never interpolate _MODULES into a tag and drop unused mut Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
bb12bc766c |
serve the parser wasm in the vitest bridge (#11420)
* fix(ai_evals): serve the parser wasm in the vitest bridge The parsers behind inferArgs load their wasm from a vite ?url path that only a dev server answers, so every script draft in the global evals got an empty schema. test_run_script then told the model its arguments were undeclared, and gemini rewrote a correct script until it hit max turns. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix(ai_evals): match wasm urls past a base path and decode them Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix(ai_evals): decode wasm urls after the suffix check; fail the run when inference breaks Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
 Diego ImbertandClaude Opus 5.5
|
83b8954ae2 |
fix: turn the default raw app into a feature tour (#11417)
* feat: turn the default raw app into a feature tour Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: don't present ctx fields as unforgeable in the starter app Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * chore: format raw app templates Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
 
|
05bcc361af |
stop billing service accounts twice after a session refresh (#11408)
* fix: stop billing service accounts twice after a session refresh Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: refuse refresh for expired, swept or impersonation tokens Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * chore: update ee-repo-ref to 5a2b6b8527250bd12cf856d8a667a9ef3106ec60 This commit updates the EE repository reference after PR #835 was merged in windmill-ee-private. Previous ee-repo-ref: 8cc94ec6aeb603b6f6ebbe1fa95b32fbec074b74 New ee-repo-ref: 5a2b6b8527250bd12cf856d8a667a9ef3106ec60 Automated by sync-ee-ref workflow. --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com> |
||
|
|
d1260d7c5a |
feat: refuse OIDC tokens to previews that impersonate a path their user cannot write (#11396)
* feat: add provenance claims to job OIDC tokens and mark preview sub Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: require a flow or script job's version to belong to its path for deployed Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: derive app script paths server-side and test job provenance in CE Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: count an app script as deployed only when a deployed app run stamped it Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * docs: state the deployed condition for the preview sub prefix Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * feat: keep the plain OIDC sub for previews by users who can write the path Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * feat: refuse OIDC tokens to previews by users who cannot write the path Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * refactor: keep OIDC token issuance unchanged, leaving provenance to the claims Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * refactor: name the root job's trigger claim root_trigger_kind Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * feat: report the request-supplied paths in a job's provenance Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * feat: count a worker-pushed preview under deployed code as deployed Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: drop the claimed parent of a restarted flow preview Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: only read a job's modules from its args when it is a preview Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: never count a preview whose modules came in its args as deployed Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * chore: update ee-repo-ref to dba91044a174f8c9275fefea2619dc966b0f9250 This commit updates the EE repository reference after PR #832 was merged in windmill-ee-private. Previous ee-repo-ref: a703cb17776d0858255fbfd817dce132377d3e80 New ee-repo-ref: dba91044a174f8c9275fefea2619dc966b0f9250 Automated by sync-ee-ref workflow. --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com> |
||
|
Diego ImbertandClaude Opus 5.5
|
50ffad52d7 |
fix: surface raw app build errors to the session AI (#11415)
* fix: surface raw app build errors to the session AI Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: route app build state per preview and report pending builds Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: name open previews when app logs are ambiguous, re-wait on retry Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
c1b59f70dd |
feat(ai-agent): add compaction memory that summarizes older context (#10928)
* feat(ai-agent): add autocompacted memory that summarizes older context Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): compact on final-answer turns and count what a turn appended Address the pre-push review findings on the compaction path: - A turn the model answers without a tool call left the agent loop on its first iteration, so a chat-shaped step never compacted and reloaded the whole conversation on every later turn. Compaction now also runs after the loop. - The trigger measured only the last request, so a single large tool result could carry the next one past the window without ever crossing 80%. - The summarization call re-sent the usage-tracking request shape on endpoints the loop had already learned to drop it for. - The flat 8000-token summary reserve swallowed the whole target on a small context window, leaving one message in the tail and summarizing the rest. - A response cut off inside the <analysis> scratchpad was accepted as a summary. - The chat-mode memory default was a shared object the step form edited in place. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): keep Anthropic prompt counts and compact once per response Address the first CI review round on the compaction path: - Anthropic's streaming parser dropped `message_start`, the only event carrying the prompt-side counts, so a native Anthropic run reported no input tokens at all and compaction fell back to a character estimate. - A loop that exits without issuing another request — a structured-output turn does — reached the post-loop pass still holding the previous measurement and compacted a second time, or retried a failure with nothing changed. - The summarization call inherited the step's `max_completion_tokens`; a low one truncates the summary inside its scratchpad, which counts as a failure and disables compaction after three of them. - A fired trigger that found nothing to summarize said nothing. - Memory already over the window — a lowered `context_window`, or a step moved over from `auto` — had no way back, since compaction only ran after an accepted request. It now also runs once before the first one. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): state the summary's own completion cap and drop the pre-flight pass - The summarization call asked for no completion cap at all, which is "uncapped" only on the OpenAI-shaped providers: Anthropic substitutes 64000, over several Claude models' output ceiling, and Bedrock leaves the model's own small default, short enough to cut the response off inside its scratchpad. It now asks for the reserve the split already set aside, raised to the step's cap when that is larger. - Compaction no longer runs before the first request. The fallbacks the loop learns from a rejection are not known that early, so on exactly the endpoints that need them the summarization was malformed by construction: it failed, spent a strike, and the first agent request still carried the oversized conversation. A memory already past the window is repaired on the turn after a request the endpoint accepts, rather than by a pass that cannot succeed there. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): ask the summary for exactly the room the split reserved The split scales its reserve down on a small window while the request asked for a flat 8000, so the two diverged below an 80k window: on a 4k/8k model the cap alone exceeded the window and every summarization was refused, and on a 20k one a full-length summary could land the conversation back over the trigger and compact its own previous summary on the next response. Both now read one `summary_reserve_tokens`. The call also no longer inherits the step's reasoning effort. Every provider counts thinking against that same budget, so a high-effort model could spend the whole reserve before writing anything and return a summary cut off inside its scratchpad; the compaction prompt asks for an `<analysis>` block, which is the reasoning this call needs. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): charge the compaction budget for tools and the system prompt The tail budget was the whole target, but a request also carries the system prompt compaction keeps and the tool definitions, which are not in the message list at all. On a small window those are most of it: a tail sized to the full target left the next request back over the trigger, compacting again every response, and the no-usage estimate missed the tool schemas entirely so it could fail to trigger at all. Both now account for them. The reserve also gains a floor. It is the summary's output cap as well as the room the split leaves, and scaled down without one a small window gave a structured nine-section summary a few hundred tokens — truncated inside its scratchpad every time, which is discarded, which switches the mode off after three. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): count Gemini's tool-use prompt tokens in an agent step's usage Gemini splits a tool-using turn's input across `promptTokenCount` and a disjoint `toolUsePromptTokenCount`, and its thinking apart from `candidatesTokenCount`. The agent step's parser read only the headline fields, so every tool-using turn under-reported both — and the compaction trigger, which runs off the reported prompt, could not see the tool results that grew it. It now goes through the same helpers the proxy path already used. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): calibrate the compaction estimate against the measured prompt Two rounds running, the finding was "the character estimate cannot see input X" — tool schemas, then S3 attachments, which are short paths in the message list and whole images by the time a provider counts them. Enumerating those is a list that only grows, so the estimate is now scaled to the one number that is ground truth: what the provider charged for the last request. Attachments, tokenizer drift and whatever comes next fall out of that, because the estimate is only ever used relative to itself. Also stop the Gemini helpers turning an absent count into `Some(0)`. Downstream, absent means "fall back to estimating the conversation" while zero reads as an empty prompt and would hold the trigger below its threshold for the whole run. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): charge attachments what they cost and let a heavy short prefix compact The calibration conserved the conversation's total cost but spread it by character count, so an attachment — a short S3 path in the message list, a whole image or PDF once a provider expands it — was charged to the text messages around it and stayed nearly free in the split. It now carries a nominal cost of its own, which the calibration corrects a residual on rather than the whole gap. The four-message minimum also refused exactly the case that fix is for: an attachment arriving on the first or second turn can pass the trigger before four removable messages exist, and summarizing even one of them saves most of the prompt. A prefix worth a quarter of the window is now enough on its own. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): never summarize a prefix holding only a previous summary The message-count floor was carrying a second job: a fresh summary sits in a one or two message prefix, so requiring four declined it. The share threshold added last commit admits it, and a summary is reserve-sized by construction — so the post-compaction shape could spend one summarization per response swapping a summary for another the same size, shrinking nothing and losing fidelity each time. A previous summary no longer counts towards that threshold. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): take the context window from the model and drop the estimate calibration Brings compaction in line with how the AI session does the same job, which had already answered these three questions. - The window is looked up from the model. `MODEL_CONTEXT_WINDOWS` in `windmill-ai/src/model_context.rs` mirrors the session's table in `copilot/modelConfig.ts`, entry for entry and with the same matching rules; each side points at the other, since a model added to one and not the other compacts at two different sizes. A step's `context_window` becomes the override for what the lookup cannot serve, and chat mode writes none. - Provider usage is normalized where the provider's quirk is, not at the consumer. `TokenUsage::with_cache_beside_input` raises `input_tokens` to the whole prompt for Anthropic and Bedrock, which report their cached prefix beside it; the OpenAI shape already counts it inside. `prompt_tokens()` is then just `input_tokens`, rather than inferring the shape from whether a write count is present. - The estimator is no longer calibrated against the measured prompt. The session uses the provider's count when it has one and a chars/4 estimate otherwise, with nothing in between, and a tail sized a little wrong only compacts again a turn later. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * feat(ai-agent): summarize memory down to what the database can store Without an instance object store, memory is a 100KB database row cut from its oldest message, the summary included, so compaction on a mainstream model never got to keep anything across runs. A step that persists there now runs its post-loop compaction pass against the smaller of the model's window and the cap at chars/4, about 25k tokens: the loop keeps the whole window, and what is written is a summary plus a tail that fits. The run logs when that pass summarizes, and how many messages the write dropped when one still overshoots. The editor's storage warning on the option is removed: nothing exposes the instance storage to it, so it keyed on the workspace S3 setting, which is unrelated to where memory goes. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): get a complete, billed summary out of every provider Compaction against the real providers turned up four things the stub could not: Gemini and OpenAI's reasoning models think by default and bill it against the same cap the summary must fit in, so the summarization request now asks them for their least (none, low); an OpenAI Responses call that hits max_output_tokens ends in response.incomplete, whose usage the parser dropped, so that summarization went unbilled; a summary that quotes </summary> when it describes its own instruction was cut off at the quote, on the agent step and the AI session alike; and the prefix could end on an unanswered user message, after which the instruction reads as part of that turn (Anthropic merges the two outright). The tail now starts on a user message, and both prompts tell the model the instruction is not part of the conversation. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): compact down to half the window, on the agent step and the AI session The gap between the 80% trigger and the target is what one compaction buys, and every summarization request carries most of the window. At a 70% target a 128k model summarized about 13k tokens of prefix for a summary of up to 8k, so each ~100k-token request bought a few turns of room before the next one re-summarized the previous summary. At 50% the same request frees about 30k. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): drop the workspace-S3 memory hint and state the database bound in the tooltip The memory field warned that memory is kept in the database whenever the workspace had no S3 storage. That setting has no bearing on where memory goes: the instance object store decides, and nothing exposes it to the editor. The field's tooltip now describes both memory kinds and states the database bound unconditionally; the run log says what happened. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): send the summarizer its tool history as text The summarization request carries no tool definitions, and Bedrock's Converse API rejects toolUse/toolResult blocks that arrive without them, so on Bedrock every summarization of a prefix holding a tool call failed silently until the breaker tripped. The prefix's tool calls and results now reach the summarizer rendered as text, on the agent step and in the AI session's compaction, which goes through the same proxy. Also drops the TokenUsage::prompt_tokens accessor, which had become a plain read of the normalized input_tokens, and shortens the context window field's description. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): price attachments from the provider count, bound storage in bytes, effort per pro model Addresses two Codex rounds and a leftovers audit. - Attachments were priced at a flat 1500 tokens in the split, so a multi-page PDF (tens of thousands of tokens to the provider, a short S3 path in the message list) could be kept in the tail or leave no prefix worth summarizing. They are now priced from the provider's count for the request that carried them, less that request's text, with the 1500 floor where nothing was counted. - The database storage bound measured the provider's token count, but the 100KB cap is bytes and repetitive text packs several characters per token. The persist pass now measures the serialized conversation. - The summarizer forced `low` on every reasoning model, which the pro variants reject (gpt-5-pro takes only high, gpt-5.2-pro starts at medium); they now get no effort. - Dropped the unused prompt_tokens accessor and its orphaned assert, an unused PartialEq, a needlessly public lookup, and fully-qualified Gemini calls; refreshed stale comments and the memory_id schema doc; regenerated the flow schema artifacts. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): evict a heavy attachment into the summarized prefix, not the tail Pricing attachments from the provider count was not enough on its own: a leading attachment is a user message, and the boundary rule pulled the last unanswered user turn back into the kept tail to keep it with its answer. For a heavy attachment that dragged it into the tail — or, at the front, emptied the prefix — so it was never summarized and rode every request. The boundary now moves forward instead, keeping that user turn and its answer in the summarized prefix. Verified on the running instance: a 25k-token PDF on a 30k window is summarized out on the turn it overflows, and later turns drop from 26k to ~1.5k tokens. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): keep the forward boundary move off tool results and the prefix start The forward move that keeps an unanswered user turn out of the tail had two edges the third Codex round found: advancing past the user could land the boundary on a tool result (its tool_calls then summarized away, orphaning it), and with no system prompt the summarizable prefix starts at 0, so a trigger firing while the tail estimate fit everything indexed below the start and panicked the task. The forward scan now skips tool-opening boundaries, and the move is guarded above the prefix start. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): drop the step temperature from the summary request OpenAI's reasoning models (gpt-5-mini, gpt-5.1, gpt-5.2) reject `temperature` alongside any reasoning effort but their own default, so a step configured with a temperature made every summarization fail once the summarizer forced a low effort — history then grew unchecked. The internal summary call now omits the step's temperature: a structured extraction does not need a set one, and omitting it sidesteps each provider's temperature-versus-reasoning rules. Confirmed against the API that low + temperature is refused on those models. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): compact an oversized loaded memory before the first request Compaction was reactive, taken only after a request the endpoint accepted, so the fallbacks the loop learns from a rejection are known first. But a memory loaded from an earlier run can already exceed this run's window — the step was switched to a smaller model, or a run under a wider one persisted more than fits — and that first request then overflows and fails the run, with every retry reloading the same history and failing again. A pass is now taken up front, off the character estimate, before the first request. It uses the default request shape; an endpoint needing a fallback may reject this one summary, which is non-fatal, and mainstream providers need none. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): under the storage bound, trigger on the max of bytes and model tokens The storage-bound pass measured only the serialized row size, so an attachment — a few bytes as an S3 path but nearly the whole model context — read as tiny and the pass skipped a compaction the model needed. It now takes the larger of the byte measure and the model's token count, since repetitive text is few tokens but many bytes and an attachment is the reverse; either being over must fire a pass. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): drop oldest turns when a summary cannot fit the window, as the AI session does An oversized loaded memory (a step switched to a smaller model, or an object-store run that persisted more than a later model's window holds) left a prefix larger than the summarizer's own window, so the summary request overflowed and failed, the memory was untouched, and every retry failed the same way. The AI session handles this by falling back from summarization to dropping the oldest turns down to the target; compaction here now does the same. When a summary cannot run — it failed, the breaker is tripped, or nothing is worth folding — the oldest turns are dropped until the conversation fits and opens on a user message, keeping the newest turn. The next request then always fits. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): drop whole turns only, keep the storage pass to bytes, refresh the count after a rewrite Three edges the seventh Codex round found, all in the drop-oldest fallback and the storage-bound measure: - drop_oldest_to_fit dropped to any point that freed enough, which could strand a tool result whose tool_calls went with the messages before it. It now drops whole turns only, always landing the boundary on a user message and never splitting the newest turn; a lone turn too big for the window is left whole rather than broken. - The storage-bound pass measured the whole model prompt against the shrunk 25k window, so a large tool roster and the system prompt — neither written to the row — tripped it on a conversation the row easily held. It measures the serialized bytes alone now; the model's own window is enforced by the in-loop passes and the pre-first-request pass, so the persisted size is all this pass is for. - A compaction rewrites the message list, so the provider's count for the request that produced it no longer lines up. The count is now cleared after any pass that rewrites the conversation, so a later pass measures the estimate over the actual messages instead of a stale, larger prompt (which could decline a summary that already fit and then drop it). The step temperature, no longer sent to the summarizer on any path, is dropped from the request struct. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): measure only the persisted messages against the storage cap Persistence strips the system prompt before writing the memory row, but the storage pass was serializing every message including it, so a large system prompt with a tiny conversation reported far over the storage trigger, and the fallback dropped the one real turn, run after run. The storage measure now serializes only the non-system messages, matching what the row actually holds. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix(ai-agent): run the model-window pass before the storage-bytes pass post-loop A turn the model answered without a tool call broke before the in-loop compaction check, so on database-backed memory its only pass was the storage one, which measures bytes. An attachment fills the model context but is a few bytes in the row, so that turn never compacted and a follow-up could overflow the model. The post-loop now runs a model-window pass first, off the provider's count, then the storage-bytes pass when the row is smaller than the model — both limits enforced for a chat-shaped step, not just the one that happens to bind. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix: simplify agent compaction and preserve execution history * fix: remove unused compaction history setting * fix: preserve answers and recover rejected agent context * refactor: make agent compaction transactional * fix: skip agent summaries that cannot fit retained context * fix: explain skipped agent context compaction * fix: retain recent agent memory when storage compaction cannot fit * fix: start retained agent memory at a user turn * fix: reject unsafe agent memory truncation on storage fallback * docs: clarify agent context window override scope * fix: keep recent turns verbatim when compaction memory outgrows storage Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix: keep the compaction summary out of the agent's answers Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * fix: shorten the agent context window help text Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01ViJyjUmidDYV2m6ifQdLeH * chore: update ee-repo-ref to 942d4013f36edac1fc9a9addbdb02198db1c7a05 This commit updates the EE repository reference after PR #812 was merged in windmill-ee-private. Previous ee-repo-ref: 8ca1682ce6106ba6ea96894fbe606dac64102eb6 New ee-repo-ref: 942d4013f36edac1fc9a9addbdb02198db1c7a05 Automated by sync-ee-ref workflow. --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com> Co-authored-by: Ruben Fiszel <ruben@windmill.dev> |
||
|
hugocasaandClaude Opus 5
|
cf5c49c3dc |
feat: restart perpetual runs on the version a deploy makes runnable (#11200)
* feat: opt-in move of perpetual runs to a newly deployed script version Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: keep the perpetual-run opt-in across relocks and check the new version's tag Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * test: pin the tag check on a perpetual version switch Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: count perpetual runs past the first queue page in the deploy prompt Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: restart perpetual runs on the version a deploy makes runnable Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: claim a perpetual run and queue its replacement in one transaction Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: honour a cancel that lands after the worker last read its queue row Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * refactor: leave the lost-cancel fix to its own PR and match the scale down to 0 wording Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: run a preprocessor the deployed version adds over the arguments carried over Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * style: shorten the wording of the modal's argument warning Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * refactor: never preprocess a restarted perpetual run, as every other restart does Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: preprocess for a replacement whose run had not been through one Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: push a perpetual replacement without the deployed debounce settings Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: name the runs the deploy button restarts Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * perf: skip the perpetual restart lookups on a deploy that is not perpetual Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: move perpetual runs before anything that can fail after the deploy commits Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: report cancellation on the queue listing so the deploy prompt can skip it Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: pass the tag workspace to the availability check after the merge Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * style: say which arguments are defined differently and which values are kept Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: resolve a dynamic tag before checking it for a restarted run Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * test: pin that a deployed dynamic tag is checked as it resolves Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
|
bda00c6cad |
fix: run custom schedule handlers as the schedule (#11398)
* fix: run custom schedule handlers as the schedule Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: correct handler identity comment and pin preset path check Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: pin preset handler paths instead of matching their name Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: warn on unpinned preset-named schedule handlers Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * chore: update ee-repo-ref to 12a83524b81368043bc939caeb3764fd27a32523 This commit updates the EE repository reference after PR #833 was merged in windmill-ee-private. Previous ee-repo-ref: bcfb0838f105d5bffbe49c6ed787116662ebe796 New ee-repo-ref: 12a83524b81368043bc939caeb3764fd27a32523 Automated by sync-ee-ref workflow. --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
6064aecdec |
feat: stop cross-origin isolating the raw app editor (#11411)
* feat: stop cross-origin isolating the raw app editor Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0166irwM15U8vZjc58MsU54U * docs: drop stale raw app editor reload comments Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0166irwM15U8vZjc58MsU54U * feat: drop wm_coep from the default raw app embed snippet Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0166irwM15U8vZjc58MsU54U * chore: pin the UI builder that type-checks without cross-origin isolation Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0166irwM15U8vZjc58MsU54U * docs: name the proxy-isolated case behind the remaining isolation checks Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0166irwM15U8vZjc58MsU54U --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Diego ImbertandClaude Opus 5.5
|
2f87bc6a2d |
fix: keep new-menu submenus open while crossing the gutter (#11414)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
e6df8d78d4 |
fix: bound a resumed session fork's wait, keep its intent while in flight (#11413)
* fix: bound a resumed session fork's wait, keep its intent while the fork is in flight Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: wait for a session fork another request is creating instead of dropping it When the fork is refused as already being created by a creation whose id was lost, the session waits for it to show up among the user's workspaces and adopts it, rather than aborting the send. An 'already exists' answer is adopted like a duplicate key. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Diego ImbertandClaude Opus 5.5
|
da5c58de2a |
fix: keep secret variable values hidden when toggling secret (#11383)
* fix: keep secret variable values hidden when toggling secret off and on Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: clarify the secret unlock hint and hide it from read-only users Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: name the secret toggle and skip the load lock on draft-only variables Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: name the secret toggle with aria-label so its heading does not flip it Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: re-hide a cleared secret value when turning secret back on Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
054109d855 |
fix: create workspace forks in the background, with progress (#11406)
* fix: create a fork in the background so a proxy timeout cannot cut it create_fork copies the whole workspace inside the request, which can run past the route timeout of an ingress in front of Windmill (Envoy's 15s default), and the UI then reports a failure for a fork still being made. create_fork?background=true now returns once the request is validated and records the copy in workspace_fork_creation, which the new fork_creation_status endpoint reads. The wizard and AI-session forks use it. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: drain background forks on shutdown and fork in the background from the CLI Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: settle fork polls by the fork's existence, adopt in-flight creations Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * feat: show which part of the copy a fork being created is in The background copy reports its phase (data tables, settings, resources, scripts, flows, apps, drafts, triggers) through a watch channel. The heartbeat task records it on the fork's creation row as soon as it changes, and fork_creation_status returns it. The fork wizard shows it on its button, and the CLI logs each step. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: join a retried fork creation server-side, settle status by the fork's existence A retry from the same user and parent joins the creation in flight instead of being refused, so clients no longer match the refusal's wording, and another requester can never adopt it. The status route reports a fork that exists under its parent as completed, whatever its run's record says. Clients give up on failing polls after a time window rather than a count, which a rolling deploy can exhaust. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: drop fork-creation joins and existence probes A second request for a fork being created is refused again; only the AI-session fork, whose request never varies, waits for its own earlier one. Clients recognise a server without background forks by its synchronous answer instead of probing for a workspace by id, which could name another parent's fork. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: poll a background fork by the id of its own attempt create_fork?background=true answers with a creation id, and the status route reads that attempt only, for the user who started it. A retry that reuses the fork id is a new attempt, so a poller never reads another attempt's outcome. The AI-session fork no longer adopts a creation in flight, which it could not tell apart from someone else's. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: record a fork's completion in its own commit, resume a session's fork after reload The attempt is marked complete in the transaction that creates the fork, so the status never infers completion from a workspace that may belong to another request. An AI session keeps the creation id on its pending fork and, after a reload mid-copy, waits for that attempt instead of requesting the fork again. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben Fiszel
|
3936191f46 |
Revert "feat: add pull_batch to claim jobs for many waiting workers at once (…" (#11412)
This reverts commit
|
||
|
Ruben FiszelandClaude Opus 5.5
|
9f40cdca62 |
feat: add pull_batch to claim jobs for many waiting workers at once (#11350)
* feat: add pull_batch to claim jobs for many waiting workers at once Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: keep jobs admitted by earlier batch passes when a re-pull fails Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: keep suspended flows first on every batch re-pull pass Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
ede4103b00 |
feat: share AI guidance with the CLI skills and lint flow groups (#11397)
* feat: lint AI agent tool names and flow groups in wmill lint * refactor: assemble chat and CLI AI guidance from one topic table * feat: share flow groups, reuse and pipeline guidance with the CLI skills * feat: share raw app, data table and secret guidance between chat and CLI * docs: document the shared AI guidance source for contributors * fix: keep wmill lint running on flows with malformed collections * fix: reject skill descriptions that are not plain YAML text * fix: tighten fence typos, script base scope and app prompt order * fix: skip tool name checks on agent steps linked to a saved agent * fix: catch any misspelled prompt fence and soften the tool name claim * fix: align cli eval harness with the files and steps wmill init adds * fix: drop cli eval checks that expect unrequested deploy commands * fix: list ansible as mainless and c# Main in script base guidance |
||
 GuilhemandClaude Opus 5.5
|
d44c901647 |
replace the AI sessions beta banner with a feedback link (#11401)
* feat: make the AI sessions beta banner dismissible Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * feat: move AI sessions feedback link into assistant settings Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * refactor: reduce the sessions beta gate setter to opt-in only Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: size the sessions activate buttons with unifiedSize Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: keep the sessions page activate button at its previous height Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
AlexRV12
|
c34abf7330 | fix: refuse flow preview restarts from runs the caller cannot read (#11407) | ||
|
|
8741843d2e |
feat: external instance cluster for data tables and Ducklake catalogs (#11197)
* fix pg_dump stuck on version 17 on nix * fix(datatables): refuse a malformed role annotation instead of ignoring it `-- Role operator`, `-- role operator;` and `-- role operator -- why` all failed the annotation parser's exact-match rule, so the query fell through to the data table's default role and ran, silently, under a login the author did not choose. Naming a role exists precisely to not do that. A leading comment whose first word is `role` is now an annotation attempt: the keyword matches case-insensitively, one trailing `;` is tolerated, and anything else is an error naming the line. Only callers that already know the target is a `datatable://` reference ever run this, so ordinary SQL keeps its comments. Also bumps the dev shell's postgres client to 18 — it trailed the server the dev database runs, which takes out every data table export, clone and fork-with-data. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012ti5HyeTikPMYyW8YSdiHR * fix(datatables): refuse a malformed role query string instead of ignoring it `?Role=analytics`, `?role=` and `?x=1&role=…` all fell through the reference parser's exact-match rule, so the connection resolved to the data table's default role and ran under a login the caller never asked for — the URI half of the same trap as a malformed `-- role` annotation. The key now matches case-insensitively, and anything else in the query string is an error naming it; `role` is the only parameter a reference takes. Callers that only need the entry keep a lenient `datatable_ref_name`, since they never act on the role. The DuckDB `ATTACH` parser propagates it rather than attaching under the default. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012ti5HyeTikPMYyW8YSdiHR * fix(datatables): carry the role annotation into the row_to_json retry The retry rebuilds its SQL from `pruneComments(code)`, so the leading comment block never reached the second attempt — and with it the `-- role <name>` line that decides which login the query runs as. The retry connected as the data table's default role instead, so a query the first attempt was denied could succeed on the second, reported as "recovered with the row_to_json fix". Carry the leading comment block over. The retry itself is unchanged. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BjfMkJyKzodxkobqGZ6Lqb * chore(datatables): don't mount the roles UI until the ACL editor lands Enforcement ships first. The permissions drawer is what turns roles on, and the catalog section is what creates them — both are only useful once there is a way to grant a role the privileges it needs, which arrives with the ACL editor. Left mounted they would offer a feature whose other half does not exist. The two components are complete and reviewed; only their call sites here are commented out, with a note pointing the follow-up PRs at them. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BjfMkJyKzodxkobqGZ6Lqb * fix(datatables): honour `-- role: x`, and fix the DuckDB attach test Two review findings, both real. `attach_datatable_parses_name_and_role` never compiled: `parse_attach_datatable` returns `Result<Option<_>>` now and one call site kept a single `unwrap`. Its `?Role=analytics` case also asserted a refusal, contradicting the parser in the same commit, which matches the key case-insensitively. Replaced with the cases that are genuinely malformed, and a positive one for the cased key. `-- role: analytics` fell through to the default role — the silent fallback the strict parser exists to remove, for the spelling most likely to be typed. The keyword now accepts an optional colon, attached or spaced, while a word that merely starts with it (`rolebased`) is still not an attempt. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BjfMkJyKzodxkobqGZ6Lqb * fix(datatables): clone a fork's pointer instead of failing after the copy Forking a fork with cloning left an orphan database. The preflight resolves the pointer and sees the governing entry, so both endpoints ran and filled the new database; `apply_forked_datatable` then refused the inherited pointer and rolled the fork back, stranding a registered `wm_fork_*` that no entry names and whose name blocks the retry. Refusing earlier would have been the smaller change, but forking a fork and cloning worked before pointers existed, so it would trade an orphan for a regression. Resolve what the pointer names and write the terminal entry the clone needs: the whole `database` object rather than a patch of its `resource_path`, since a pointer has none, and `reference` removed with it. Also accepts `-- role=x` and `-- Role = x`, two more spellings that fell through to the default role. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BjfMkJyKzodxkobqGZ6Lqb * fix(datatables): refuse to roll back the catalog while roles exist The down migration dropped the table and left every role behind: live Postgres logins whose passwords only that table carried, so after a revert Windmill could neither use, disable nor delete them, and re-applying could not recreate them because the names were taken. Cleaning up here is not possible either — dropping a role means reassigning what it owns in every instance database, and a migration runs in one — so it now refuses while the catalog is non-empty and says to delete the roles through instance settings, which does the cluster work. Also enforces the instance-only invariant the resolved-pointer clone relies on rather than only asserting it in a comment. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BjfMkJyKzodxkobqGZ6Lqb * refactor(datatables): settle clonability in one place, before anything is created A clone is three stages a workspace apart — `create_pg_database`, then `import_pg_database`, then `apply_forked_datatable` inside the fork transaction. Only the third can roll back, and `CREATE DATABASE` is not transactional, so any refusal that lives there strands a registered `wm_fork_*` that no entry names and whose name blocks the retry. That orphan has now been fixed three times, most recently reintroduced by a guard added one commit ago. Patching each new refusal into the first endpoint is not the fix; having two places that can refuse is. `ensure_datatable_is_clonable` now answers every reason a copy can be refused and returns what it resolved, and the stage that writes the entry only does the work. Also takes an ACCESS EXCLUSIVE lock before the rollback guard counts, so a role created concurrently cannot slip between the check and the drop. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BjfMkJyKzodxkobqGZ6Lqb * fix(datatables): let a retried clone reclaim its own leftover database A clone creates its target database one request before it copies into it, and the fork that would name it is written a request after that. Any failure in between — a pg_dump error, a bad restore, a dropped connection, the source's roles changing mid-flow — left a registered `wm_fork_*` that no entry names, and every retry then failed on its name. This predates data table roles. `create_pg_database` now reclaims such a leftover before creating: only a `wm_fork_*` database Windmill registered as a data table database and that no data table or ducklake entry names, in any workspace, archived ones included. The drop never terminates connections, so a clone still copying into it makes the reclaim fail instead of being cut off. It is limited to callers who administer the source — reaching it is not enough, since on a data table without roles every member reaches it — and anyone else gets the refusal an existing database always got. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * Revert "fix(datatables): let a retried clone reclaim its own leftover database" This reverts commit |
||
|
hugocasaandClaude Opus 5.5
|
797147ea7a |
fix: report a worker's last job when it ran under one poll interval (#11399)
* fix: report a worker's last job when it ran under one poll interval Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: share the unreported job slot with the interactive worker shell Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * test: pin that a main-loop ping without a job keeps the last one Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
hugocasaandClaude Opus 5.5
|
97fa55b719 |
feat: detect and alert when a schedule skips occurrences (#10917)
* docs: plan for detecting skipped schedule occurrences Design plan only, no implementation. Records the scheduler's re-anchoring behaviour, the measurements behind it, and the three-piece design that came out of reviewing the alternatives. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * docs: state the user-facing outcome in the schedule plan The plan described the mechanism but never what a user would see, which made it hard to judge what the work is worth. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * docs: state which cause the schedule plan catches, and correct its scope Records which of the two causes each piece covers, and corrects the overrun scope: a script schedule carrying retry or dynamic_skip is pushed as a SingleStepFlow, so it re-arms at step 0 entry and its occurrences overlap like a flow's. Resolves the no_flow_overlap question, splits the read-time work into bounded detection and editor-only counting behind measured croner costs, and fixes the delivery order. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * feat: count the occurrences a schedule skipped A schedule that overruns its interval, or waits for a worker, silently loses the occurrences in between: the scheduler keeps one queued occurrence and re-anchors on the clock, so nothing records that a run was due and never happened. Recovers the sequence from rows that already exist rather than writing per occurrence. `push_scheduled_job` anchors on `now_from_db` inside the transaction that inserts the job, and `v2_job.created_at` defaults to that same transaction timestamp, so `scheduled_for = find_next(created_at)` holds exactly and the whole occurrence history is derivable. The schedules list reports how many of the recent runs were followed by a lost occurrence, and a new occurrences endpoint carries the per-run wait and duration behind it. Detection is one `find_next` per gap, which stays bounded on a full page; counting walks the gap and runs only for a single schedule. The one write is `occurrence_baseline_at`, advanced at create, edit, re-enable and re-arm. Gaps older than it span a pause, a cron change, a re-enable or a reconciler re-arm, none of which mean runs were lost. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * feat: show the wait and run time behind a schedule's skipped occurrences The list badge says a schedule is losing runs; this says which of the two causes did it. A large wait means not enough workers, a long run means the job outgrew its interval, and the pair is what tells them apart. Sits under the existing upcoming-events panel, so due and overdue read together. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * feat: flag a schedule that is running late right now Reconstruction is retrospective: a gap only appears once the next occurrence has a row, which needs the current one to finish. A schedule wedged mid-run shows nothing until it moves, which is the case an operator most wants to see. An occurrence still in flight past the time its own successor was due will cost that successor, so `now > find_next(scheduled_for)` is the signal, needing no threshold and self-calibrating across a daily and a per-minute schedule. It applies only where occurrences serialize; an overlapping schedule starts its successor on time and would flag constantly while healthy. The queue is read in one aggregating pass keyed on (trigger, runnable_path) rather than a subquery per schedule, and an overlapping schedule holds more than one root row, hence the aggregate. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * feat: run the schedule overrun alert from the monitor pass Wires `schedule_overrun_alerts` in next to `jobs_waiting_alerts`, every 30 iterations (~5 min). Its Enterprise implementation lives in windmill-labs/windmill-ee-private#772; only the wiring and the OSS stub are here. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * chore: refresh the sqlx offline cache Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * feat: record and alert when a schedule skips occurrences push_scheduled_job compares each chained occurrence with the slot after the previous one. A gap is written to schedule.skipped_occurrences off the push transaction, alerts once when a clean schedule starts skipping, and recovers on the next clean chain. The schedules list shows a badge. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * feat: alert only on a streak of skipping runs, keep a recent skip visible The skip state now describes the current streak and is written in the push transaction, so it commits or rolls back with the push. The alert fires once when 3 runs in a row skipped, and the list keeps a muted badge for 7 days after the latest skip. Editing or toggling a schedule resets it. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * fix: name the missed-occurrence state after what it counts, alert only once committed Renames the columns to late_run_streak, missed_occurrences and last_missed_at, keeps the missed count after a streak ends so the muted badge can show it, and rewords both badges. The alert task now reads the streak FOR SHARE, which waits for the push transaction, so a push that rolls back and retries alerts once. A failed slot count leaves the streak untouched, and a schedule deleted mid-push no longer fails it. Adds an integration test for the streak and its reset. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * fix: recover the late run alert, store the missed slot, name it missed throughout The alert now recovers (and so acknowledges itself) when a streak that alerted ends on a run on time, under the schedule:{path} resource used by the other trigger alerts. last_missed_at records the last missed cron slot rather than when the late run chained, and the counting helpers say missed, since skipped already names occurrences queued and not run. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * fix: scope the late run alert to its workspace, acknowledge it on edit, toggle and delete Recovery acknowledges alerts by resource alone, so the resource now carries the workspace. Editing, toggling or deleting a schedule clears its streak and a disabled or deleted one never chains a run on time, so those handlers acknowledge its open alert after committing. Past the 1000-slot cap, last_missed_at falls back to the detection time. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 * refactor: raise the late run alert like the other critical alerts Drops the recovery, the workspace-scoped resource and the acknowledgement on edit, toggle and delete: the alert now fires once per streak with no resource and is acknowledged from the alerts feed, as the trigger and job failure alerts are. The FOR SHARE read stays, so a push that rolls back across the flow path's retries still alerts once. Notes in openapi that past 1000 misses in one late run the count is a floor and the time approximate. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LJ9wpjWp2YgLUSqt1Ai5d6 --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
|
17448c97d3 |
count trigger suspend, resume and discard (#11405)
* feat(telemetry): count trigger suspend, resume and discard Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor: shorten the telemetry disclosure to one line per category Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * docs: correct the resume branch comments and note the pre-commit fire count Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: name feature adoption in the telemetry disclosure Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * chore: update ee-repo-ref to 9855e1b7a43a0a33e04f8accf1c497af3fd9b139 This commit updates the EE repository reference after PR #834 was merged in windmill-ee-private. Previous ee-repo-ref: 1d5b128ec956c156fe549cf099ba0dbc1b6467bf New ee-repo-ref: 9855e1b7a43a0a33e04f8accf1c497af3fd9b139 Automated by sync-ee-ref workflow. --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com> |
||
|
GuilhemandClaude Opus 5
|
f367eaf6d0 |
feat: run turns in several flow chat conversations at once (#11202)
* feat: run turns in several flow chat conversations at once Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep finished turns finished and cached chats current in the flow chat pool Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: attribute a turn's rows by job id as well as sequence Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: count only real stream updates and retry the job-id read Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep a chat that holds an unsent draft, and take one back when its first turn is withdrawn Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: ignore a stale running-turn snapshot, keep a withdrawn chat's draft, poll after clean stream ends Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: follow the turn running now when the listing named one already over Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep replacement turns and SSE fallback moving * fix: keep replacement turn handoffs active * fix: preserve unread badge line height * fix: settle local fallback handoffs * fix: settle refused turn handoffs * fix: scope turn handoffs to conversation * fix: drop stale turn handoffs * refactor: move the queued message and 409 handling into per-conversation turns Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address cubic's review of the parallel flow chat turns Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: clear a stale failure on refresh, and tighten the docs and test waits Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: recover running rows past the first page, and drop the failure a re-read disproves Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * refactor: check the running-turn query at compile time, and narrow what a refresh clears Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: settle a failed turn only from an answer that turn wrote Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: settle a failed turn from its own answer, and only while it is still the failure shown Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: drop a failure whose answer arrived even when a newer turn owns the error Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: free an answered failure whatever the turn that started meanwhile is doing Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: drop a rows read that a turn outran, rather than merging it under newer messages Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: drop a rows read whose conversation was left and opened again Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: hand over a file still being read when its composer goes Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: count a drop's routing as work in flight, so its file is handed over too Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: hold the send until every file a conversation is owed has landed Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * refactor: keep a panel mounted per conversation instead of handing its draft over Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep the withdrawn chat whose composer was written in, not the empty one Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep the chat in front of the reader when both withdrawn composers were written in Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep a retry's own run arguments when a turn elsewhere refuses it Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: name panels apart across pools, and read a flow's inputs when its chat is built Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
GuilhemandClaude Opus 5.5
|
e68ff0d969 |
feat: add tree view to the schedules page (#11360)
* feat: add tree view to the schedules page Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: keep schedule job previews visible inside tree folders Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: keep job preview loading while hovered and close it on scroll Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: keep schedules outside u/ and f/ in the tree view Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * feat: add tree view to the trigger list pages (#11400) * feat: add tree view to the trigger list pages Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * refactor: let a TreeViewState own the tree view setting Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: drop the doubled bottom border at the end of a tree Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
 Ruben Fiszelandrubenfiszel
|
651a6b01f5 |
chore(main): release 1.819.0 (#11364)
* chore(main): release 1.819.0 * Apply automatic changes --------- Co-authored-by: rubenfiszel <275584+rubenfiszel@users.noreply.github.com>v1.819.0 |
||
|
Ruben FiszelandClaude Opus 5.5
|
cedd6dc901 |
fix: hold interpolated references and captures to the token path scopes (#11391)
* fix: hold interpolated references and captures to the token path scopes Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: let a resource read cover its own linked secret variable Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: resolve policy-granted app upload resources on the viewer's rls Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: cover multi-secret linked variables and keep capture paths out of refusals Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
163a4ffa4e |
fix: scope flow resume to its workspace and minting to the job's run (#11392)
* fix: scope flow resume to its workspace and minting to the job's run Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * docs: name the lineage columns resume minting checks Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
14a2619ad2 |
fix: gate batch rerun on job read access, scope started_at to workspace (#11387)
* fix: gate batch rerun on job read access and scope started_at lookup to the workspace Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * test: assert batch rerun denial comes from the read gate Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
3eaf2888c0 |
fix: scope workspace dependencies create to the path workspace (#11385)
* fix: scope workspace dependencies create to the path workspace Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * docs: state the workspace_id must-match contract in the spec and struct Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
AlexRV12andClaude Opus 5.5
|
5e59cefd1f |
fix(frontend): apply operator write locks from the session's operating workspace (#11395)
Claude-Session: https://claude.ai/code/session_01VAj4mmm2YThZLVkivrgsbb Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
f4dcaf3e45 |
fix: list only the paths the caller can read in path autocomplete (#11388)
* fix: list only the paths the caller can read in path autocomplete Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: bound the path autocomplete cache by total path count Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
ec6ec1b06f |
fix: judge IPv4 embedded in IPv6 and pin the object storage test connect (#11389)
* fix: judge IPv4 embedded in IPv6 and pin the object storage test connect Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: let the public-only object store client reach the egress proxy Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: refuse private IP literals and the proxy host in the public-only store client Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: refuse the egress proxy as a target whether named or an IP literal Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Diego Imbert
|
22b5a1cd62 | fix: keep test panel controls off the args form in debug mode (#11382) | ||
|
Diego ImbertandClaude Opus 5.5
|
d76a962331 |
feat: add hub sync button to the resource types tab (#11375)
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Diego ImbertandClaude Opus 5.5
|
d9c7d71f07 |
fix: redesign run not found page and fix switching to the right workspace (#11374)
* fix: redesign run not found page and clear stale not-found on workspace switch Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * refactor: use design-system Button for workspace rows on run not found page Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
hugocasaandClaude Opus 5.5
|
60ef82196e |
fix: keep smtp_clicktracking_off when syncing instance config (#11372)
* fix: keep smtp_clicktracking_off when syncing instance config Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * test: name smtp regression test after what it guards Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
|
990a726409 |
feat: add provenance claims to job OIDC tokens (#11369)
* feat: add provenance claims to job OIDC tokens and mark preview sub Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: require a flow or script job's version to belong to its path for deployed Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: derive app script paths server-side and test job provenance in CE Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: count an app script as deployed only when a deployed app run stamped it Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * docs: state the deployed condition for the preview sub prefix Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * feat: keep the plain OIDC sub for previews by users who can write the path Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * feat: refuse OIDC tokens to previews by users who cannot write the path Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * refactor: keep OIDC token issuance unchanged, leaving provenance to the claims Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * refactor: name the root job's trigger claim root_trigger_kind Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * chore: update ee-repo-ref to 421cf2a8b4f98b421e93c0fc7c1c378314a66e50 This commit updates the EE repository reference after PR #831 was merged in windmill-ee-private. Previous ee-repo-ref: 7acd384875deba4b01a502e628a153b11c82eecb New ee-repo-ref: 421cf2a8b4f98b421e93c0fc7c1c378314a66e50 Automated by sync-ee-ref workflow. --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com> |
||
|
Diego ImbertandClaude Opus 5.5
|
3838cd6ee0 |
fix: stop the schedule enabled toggle from showing unsaved changes (#11390)
* fix: keep schedule enabled toggle from reading as unsaved changes Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: only fold the enabled toggle into the baseline when it is deployed Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: skip the enabled revert once the drawer moved to another schedule Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
893e64f630 |
fix: only restart a flow on a version of its own path and workspace (#11376)
* fix: only restart a flow on a version of its own path and workspace Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * test: pin cross-workspace restart version rejection Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
90f9e59321 |
fix: only let a job's own token claim run lineage (#11367)
* fix: only let a job's own token claim its lineage on the run endpoints Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: drop an unclaimable run lineage instead of refusing the run Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: only let a job's own token run its workflow-as-code tasks Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Diego Imbert
|
47525b211a |
perf: shrink the module graph that gates first paint in dev (#11373)
* perf: shrink the module graph that gates first paint in dev * docs: drop the stale synchronous-icons claim on the import card * fix: replay search opened before its modal loads, guard lazy icons * fix: only intercept search before load where the modal mounts * perf: mount app-shell modals on first open and keep monaco off the shell * perf: load the icon map on first read, not at module evaluation * fix: report stale chunks with a reload toast, guard the home page against monaco |
||
|
Ruben FiszelandClaude Opus 5.5
|
649c43e7c1 |
fix: run an AI agent tool on the worker its own tag selects (#11370)
* fix: run an AI agent tool on the worker its own tag selects Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: run a tagged agent tool inline when this worker serves its tag Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: give an inline agent tool a job token of its own Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: report a lost tool wait to the model and cancel tools on agent timeout Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
d7a61de23f |
fix: never double-process a slow canceled flow in the zombie sweep (#11368)
* fix: leave a slow canceled flow to its live worker and bound its requeues Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UzGsxKey5g3kycKwGmpKNF * fix: give a canceled zombie flow a longer grace instead of guessing its worker Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UzGsxKey5g3kycKwGmpKNF * fix: retry a canceled zombie flow's forced completion until it lands Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UzGsxKey5g3kycKwGmpKNF * fix: claim a canceled zombie flow without waiting on its runtime row Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UzGsxKey5g3kycKwGmpKNF --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
c2d8997549 |
fix: complete a canceled flow whose worker died between two steps (#11366)
* fix: complete a canceled flow whose worker died between two steps Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UzGsxKey5g3kycKwGmpKNF * fix: complete only the stranded canceled flow and let its parent process it Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UzGsxKey5g3kycKwGmpKNF * fix: requeue a stranded canceled flow for a worker to complete its cancel Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UzGsxKey5g3kycKwGmpKNF * fix: keep a requeued canceled flow's start time Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UzGsxKey5g3kycKwGmpKNF --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
65cba2dbb7 |
perf: advance a flow step with one v2_job_status update (#11357)
* perf: advance a flow step with one v2_job_status update Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * docs: state what advance_flow_status returning None means Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: keep the merged flow advance identical for rows without a status row or with a malformed status Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014FbA4shbpCnRsUxir1GEfm * docs: note the JSON null invariant behind the empty-path no-op Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014FbA4shbpCnRsUxir1GEfm --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
|
bebd762194 |
perf: complete a job in one statement on the common path (#11355)
* perf: complete a job in one statement on the common path Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: take completion locks in one order on every path Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: keep a losing zombie completion from touching its wac parent Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: leave a flow's ping alone when a step completes during its cancel Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * test: probe only this test's completion for the lock wait Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: stamp a wac child's kept duration when its completed row exists Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01TJMJSJ2bDhYh7Shoh78Yyb * perf: leave the parent ping out of completions with no flow to ping Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01TJMJSJ2bDhYh7Shoh78Yyb * docs: note that the two completion statements must stay in step Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01TJMJSJ2bDhYh7Shoh78Yyb * chore: update ee-repo-ref to 7a256cf353db7cf64a60a09fa0de7f3a8b27f626 This commit updates the EE repository reference after PR #830 was merged in windmill-ee-private. Previous ee-repo-ref: 497137acb65e521568d46f3cbe1d66359f7f87ec New ee-repo-ref: 7a256cf353db7cf64a60a09fa0de7f3a8b27f626 Automated by sync-ee-ref workflow. --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com> |
||
|
Ruben FiszelandClaude Opus 5.5
|
e2be584ca5 |
fix: let custom workspace error handlers send email with the instance SMTP (#11365)
* fix: let custom workspace error handlers send email with the instance SMTP Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> * fix: state what the error handler email allowlist guarantees Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> |
||
|
Ruben Fiszelandrubenfiszel
|
3974bbeac6 |
chore(main): release 1.818.0 (#11294)
* chore(main): release 1.818.0 * Apply automatic changes --------- Co-authored-by: rubenfiszel <275584+rubenfiszel@users.noreply.github.com>v1.818.0 |
||
|
hugocasaandClaude Opus 5.5
|
ca8a04a869 |
fix: allow results access inside nested functions in input transforms (#11358)
* fix: allow results access inside nested functions in input transforms Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: decode escaped bracket step ids and test deferred fetch errors Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: let quickjs decode bracket step ids and match quoted forms Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: prefetch results read through spread syntax Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: keep prefetched bracket literals on a single line Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: decode prefetch step literals as data and skip unparsable ones Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: run the results prefetch outside the expression scope Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix: keep the transform expression a zero-arg iife after prefetch Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> |