Skip to content

[pull] main from simstudioai:main#30

Merged
pull[bot] merged 10 commits intoedisplay:mainfrom
simstudioai:main
Mar 13, 2026
Merged

[pull] main from simstudioai:main#30
pull[bot] merged 10 commits intoedisplay:mainfrom
simstudioai:main

Conversation

@pull
Copy link

@pull pull bot commented Mar 13, 2026

See Commits and Changes for more details.


Created by pull[bot] (v2.0.0-alpha.4)

Can you help keep this open source service alive? 💖 Please sponsor : )

waleedlatif1 and others added 10 commits March 12, 2026 01:19
* fix(blocks): remap condition/router IDs when duplicating blocks

Condition and router blocks embed IDs in the format `{blockId}-{suffix}`
inside their subBlock values and edge sourceHandles. When blocks were
duplicated, these IDs were not updated to reference the new block ID,
causing duplicate handle IDs and broken edge routing.

Fixes all four duplication paths: single block duplicate, copy/paste,
workflow duplication (server-side), and workflow import.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(blocks): deep-clone subBlocks before mutating condition IDs

Shallow copy of subBlocks meant remapConditionIds could mutate the
source data (clipboard on repeated paste, or input workflowState on
import). Deep-clone subBlocks in both regenerateBlockIds and
regenerateWorkflowIds to prevent this.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(blocks): remap condition IDs in regenerateWorkflowStateIds (template use)

The template use code path was missing condition/router ID remapping,
causing broken condition blocks when creating workflows from templates.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
…imeSegments (#3534)

* fix(traces): prevent condition blocks from rendering source agent's timeSegments

Condition blocks spread their source block's entire output into their own
output. When the source is an agent, this leaked providerTiming/timeSegments
into the condition's output, causing buildTraceSpans to create "Initial
response" as a child of the condition span instead of the agent span.

Two fixes:
- Skip timeSegment child creation for condition block types in buildTraceSpans
- Filter execution metadata (providerTiming, tokens, toolCalls, model, cost)
  from condition handler's filterSourceOutput

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(traces): guard condition blocks from leaked metadata on old persisted logs

Extend isConditionBlockType guards to also skip setting span.providerTiming,
span.cost, span.tokens, and span.model for condition blocks. This ensures
old persisted logs (recorded before the filterSourceOutput fix) don't display
misleading execution metadata on condition spans.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(traces): guard toolCalls fallback path for condition blocks on old logs

The else branch that extracts toolCalls from log.output also needs a
condition block guard, otherwise old persisted logs with leaked toolCalls
from the source agent would render on the condition span.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* refactor(traces): extract isCondition to local variable for readability

Cache isConditionBlockType(log.blockType) in a local const at the top
of the forEach loop instead of calling it 6 times per iteration.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* feat(fathom): add Fathom AI Notetaker integration

* fix(fathom): address PR review feedback

- Add response.ok checks to all 5 tool transformResponse functions
- Fix include_summary default to respect explicit false (check undefined)
- Add externalId validation before URL interpolation in webhook deletion

* fix(fathom): address second round PR review feedback

- Remove redundant 204 status check in deleteFathomWebhook (204 is ok)
- Use consistent undefined-guard pattern for all include flags
- Add .catch() fallback on webhook creation JSON parse
- Change recording_id default from 0 to null to avoid misleading sentinel

* fix(fathom): add missing crm_matches to list_meetings transform and fix action_items type

- Add crm_matches pass-through in list_meetings transform (was silently dropped)
- Fix action_items type to match API schema (description, user_generated, completed, etc.)
- Add crm_matches type with contacts, companies, deals, error fields

* fix(fathom): guard against undefined webhook id on creation success

* fix(fathom): add type to nested trigger outputs and fix boolean coercion

- Add type: 'object' to recorded_by and default_summary trigger outputs
- Use val === true || val === 'true' pattern for include flag coercion
  to safely handle both boolean and string values from providerConfig

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Lakee Sivaraya <71339072+lakeesiv@users.noreply.github.com>
Co-authored-by: Vikhyath Mondreti <vikhyath@simstudio.ai>
Co-authored-by: Vikhyath Mondreti <vikhyathvikku@gmail.com>
The GET /rest/api/3/search/jql endpoint requires an explicit `fields`
parameter to return issue data. Without it, only the issue `id` is
returned with all other fields empty. This adds `fields=*all` as the
default when the user doesn't specify custom fields.

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Use explicit hyphen separator instead of relying on slice offset to
implicitly include the hyphen in the suffix, making the intent clearer.

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* feat(slack): add email field to get user and list users tools

* fix(slack): use empty string fallback for email and make type non-optional

* fix(slack): comment out users:read.email scope pending app review
@pull pull bot locked and limited conversation to collaborators Mar 13, 2026
@pull pull bot added the ⤵️ pull label Mar 13, 2026
@pull pull bot merged commit e9bdc57 into edisplay:main Mar 13, 2026
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant