ci: summarize Mantis bug proof comments

Clarify that Mantis screenshot proof belongs on the bug/fix PR and add a top summary to inline evidence comments.
This commit is contained in:
Peter Steinberger
2026-05-03 17:39:46 +01:00
committed by GitHub
parent ddae71537d
commit c40f89414c
2 changed files with 12 additions and 5 deletions

View File

@@ -14,7 +14,7 @@ on:
default: main
type: string
pr_number:
description: Optional PR number to receive the QA evidence comment
description: Optional bug or fix PR number to receive the QA evidence comment
required: false
type: string
@@ -334,6 +334,8 @@ jobs:
<!-- mantis-discord-status-reactions -->
## Mantis Discord Status Reactions QA
Summary: Mantis reran Discord status reactions against the known queued-only baseline and the candidate ref. The baseline reproduced the bug, while the candidate showed the expected queued -> thinking -> done reaction sequence.
- Scenario: \`discord-status-reactions-tool-only\`
- Run: https://github.com/${GITHUB_REPOSITORY}/actions/runs/${GITHUB_RUN_ID}
- Artifact: ${ARTIFACT_URL}

View File

@@ -346,16 +346,21 @@ after the new secret has been stored.
## GitHub Artifacts And PR Comments
Mantis workflows should upload the full evidence bundle as a short-lived Actions
artifact. When the workflow is run for a PR, it should also publish the redacted
PNG screenshots to the `qa-artifacts` branch and upsert a PR comment with inline
before/after screenshots. Raw logs, observed messages, and other bulky evidence
stay in the Actions artifact.
artifact. When the workflow is run for a bug report or fix PR, it should also
publish the redacted PNG screenshots to the `qa-artifacts` branch and upsert a
comment on that bug or fix PR with inline before/after screenshots. Do not post
the primary proof only on a generic QA automation PR. Raw logs, observed
messages, and other bulky evidence stay in the Actions artifact.
The PR comment should be short and visual:
```md
Mantis Discord Status Reactions QA
Summary: Mantis reran the reported Discord status-reaction bug against the known
bad baseline and the candidate fix. The baseline reproduced the bug, while the
candidate showed the expected queued -> thinking -> done sequence.
- Scenario: `discord-status-reactions-tool-only`
- Run: <workflow run link>
- Artifact: <artifact link>