fix(ci): migrate publish pipeline to engine::triggers::list

[sergiofilhowz]

Summary

• Update the publish registry workflow and interface collection script to call engine::triggers::list instead of the deprecated engine::trigger-types::list.
• Update build_publish_payload.py to read the new triggers response key when diffing baseline vs current trigger types for registry publishing.

Test plan

• Sanity-checked normalize_worker_interface with the new { "triggers": [...] } response shape
• Verify publish-registry workflow succeeds on a worker publish run

Made with Cursor

Summary by CodeRabbit

• Chores
  • Updated internal build and deployment automation to use updated data collection methods for registry publishing.
  • Revised registry publishing workflow to align baseline trigger information generation with current system specifications.
  • Adjusted internal build scripts to maintain consistency between data collection and processing stages.
  • Updated configuration for automated registry snapshot generation to use updated data sources.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
workers	Ready	Preview, Comment	Jun 2, 2026 6:59pm

[coderabbitai]

📝 Walkthrough

Walkthrough

This PR updates the trigger type data collection pipeline to use a new engine endpoint. The workflow baseline generation and collector script both migrate from engine::trigger-types::list to engine::triggers::list, and the payload builder is updated to extract trigger arrays from the "triggers" JSON key instead of "trigger_types".

Changes

Trigger data source migration to new engine endpoint

Layer / File(s)	Summary
Workflow baseline and collector endpoint migration .github/workflows/_publish-registry.yml, .github/scripts/collect_worker_interface.py	Baseline snapshot step and collect_trigger_types() both switch from engine::trigger-types::list to engine::triggers::list endpoint.
Payload builder JSON field normalization .github/scripts/build_publish_payload.py	Baseline filtering and trigger normalization now extract data from "triggers" JSON array key instead of "trigger_types".

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

Possibly related PRs

• iii-hq/workers#67: Adjustments to build payload and collector scripts for trigger data align with existing trigger-normalization flow that structures triggers under interface data.
• iii-hq/workers#73: Overlapping refactor of trigger-type data sourcing and normalization in same scripts with different endpoint and field choices.
• iii-hq/workers#139: Parallel changes to publish-registry workflow trigger baseline collection logic and endpoint invocation.

Suggested reviewers

• andersonleal

Poem

🐰 A whisker twitch, a registry dance,
Triggers migrate through the engine's expanse,
From old endpoints to new fields we stray,
JSON keys reshape the data's ballet,
Baseline and builder aligned once more! 🌾

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The pull request title clearly and specifically describes the main objective: migrating the CI pipeline to use engine::triggers::list instead of a deprecated endpoint.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/pipeline

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

skill-check — worker

0 verified, 14 skipped (no docs/).

Layer	Result
structure	✓
vale	✓
ai	✓
render	✓

Four for four. Nicely done.

[coderabbitai]

🧹 Nitpick comments (1)

.github/workflows/_publish-registry.yml (1)

109-116: Adjust: engine::triggers::list already matches the expected per-item schema

The repo’s engine::triggers::list introspection/parsers model includes id, trigger_request_format, and call_request_format, matching what build_publish_payload.normalize_worker_interface consumes from each item under the top-level triggers key—so the baseline diff/normalization shouldn’t silently drop triggers due to field-shape changes. The only leftover cleanup is renaming trigger-types-baseline.json (and the step title) since it snapshots trigger instances rather than trigger types.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In @.github/workflows/_publish-registry.yml around lines 109 - 116, The baseline
step is misnamed: `engine::triggers::list` already outputs items matching the
schema consumed by build_publish_payload.normalize_worker_interface (fields like
id, trigger_request_format, call_request_format), so update the workflow step
name and output file to reflect that it snapshots trigger instances, not trigger
types; specifically rename the step title from "Snapshot engine trigger types
baseline" and the output file trigger-types-baseline.json to something like
trigger-instances-baseline.json (and update any downstream references) while
leaving the `iii trigger 'engine::triggers::list' --json ...` invocation
unchanged.

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Nitpick comments:
In @.github/workflows/_publish-registry.yml:
- Around line 109-116: The baseline step is misnamed: `engine::triggers::list`
already outputs items matching the schema consumed by
build_publish_payload.normalize_worker_interface (fields like id,
trigger_request_format, call_request_format), so update the workflow step name
and output file to reflect that it snapshots trigger instances, not trigger
types; specifically rename the step title from "Snapshot engine trigger types
baseline" and the output file trigger-types-baseline.json to something like
trigger-instances-baseline.json (and update any downstream references) while
leaving the `iii trigger 'engine::triggers::list' --json ...` invocation
unchanged.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 7b10b057-640a-4d55-86be-07676b4056b9

📥 Commits

Reviewing files that changed from the base of the PR and between 945b5ab and 1de2435.

📒 Files selected for processing (3)

• .github/scripts/build_publish_payload.py
• .github/scripts/collect_worker_interface.py
• .github/workflows/_publish-registry.yml