[release] v0.100.9 + Railway deploy diagnostics#4514
Closed
mmabrouk wants to merge 24 commits into
Closed
Conversation
…ects Switching projects while on an entity-scoped page (evaluation, playground, testset) kept the old entity id in the URL. Since that entity does not exist in the target project, the user landed on an empty screen. Preserve only the top-level section segment on project switch and drop nested entity ids. All top-level sections have index pages, so the truncated path always resolves. Logic extracted to a pure helper with regression tests.
…lity table The observability table's evaluator metric columns only carry a slug, but evaluatorReferenceAtomFamily resolved names solely through its id-based workflow-revision query. Slug-only callers fell through to a minimal reference that returned name=slug, leaking slugs into the table. Add a slug-resolution branch that matches the slug against the loaded app+evaluator union list (workflowsListQueryStateAtom) to surface the real name, deferring while the list loads. Mirrors appReferenceAtomFamily.
Render archived evaluators through the same single PageLayout as the active Evaluators page, with a Back arrow + title in the standard h-11 header row. Drops the ArchivedEntityLayout wrapper, which double-framed the page (extra p-4 + tall stacked Back/title/subtitle header) and pushed the table ~60-70px lower, causing a visible vertical shift when toggling between Evaluators and Archived.
…aluator-slugs-instead-of-names
…luation-or-playground-in-url
…aluator-slugs-instead-of-names
…luation-or-playground-in-url
…ble-shows-evaluator-slugs-instead-of-names
…ts-keeps-evaluation-or-playground-in-url
[fix] Resolve audit log placement
…-table-shows-evaluator-slugs-instead-of-names [FE Fix]: Show evaluator names instead of slugs in observability table
…ge-layout-fix [FE Fix]: Align archived evaluators page layout with active page
…ts-keeps-evaluation-or-playground-in-url
…jects-keeps-evaluation-or-playground-in-url [FE Fix]: Correctly handle project changes from entity scoped pages
Failed Railway preview deploys reported only 'Process completed with exit code 1'. The real cause was either swallowed by >/dev/null in the deploy scripts or lived only in the Railway dashboard, which CI never surfaced. - Add install_error_trap to bootstrap/configure/deploy-from-images so a failure prints the command, exit code, and a file:line call stack. - railway_call now prints failures to stderr instead of stdout, so callers that send stdout to /dev/null still surface the underlying railway error. - Add dump_railway_logs: on a failed deploy, pull the tail of key services' Railway logs (Postgres, alembic, api, ...) into the job. - Persist setup/deploy output to a log file, upload it as an artifact, and write a step summary with the log tail on failure. Diagnostics only; no deploy logic changes. The artifact-upload steps are marked continue-on-error so they can never fail an otherwise-passing job. (cherry picked from commit d97cf94)
Address CodeRabbit review. The ERR handler printed $BASH_COMMAND and railway_call printed railway's output on failure. configure.sh passes real secret values as CLI args to 'railway variable set' (POSTGRES_PASSWORD, AGENTA_AUTH_KEY/CRYPT_KEY, *_API_KEY), so a failure could emit plaintext secrets into the uploaded deploy-log artifact. Add _railway_redact (masks KEY=value for PASSWORD/TOKEN/SECRET/KEY keys and scheme://user:password@host) and apply it to every diagnostic path: railway_call failure + rate-limit output, the ERR handler command, and dump_railway_logs. The success path stays unredacted so callers that parse 'variable list -k' output (e.g. resolve_postgres_password) keep working. (cherry picked from commit 63136af)
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
dosubot
Bot
added
size:L
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
📝 WalkthroughSummary by CodeRabbit
WalkthroughBumps release to 0.100.9; separates EE retention DAOs/services/routers and moves events query into EE with entitlements checks; hardens Railway scripts/workflows with redaction and persistent logs; refactors Audit Log EE wiring and Settings injection; extracts buildProjectSwitchHref with tests; and applies version bumps. ChangesRailway Infrastructure Hardening
Events and Tracing Retention Refactoring
Frontend Audit Log Architecture
Frontend Navigation Helper
Frontend Evaluators UI Cleanup
Version Bumps and Miscellaneous
🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly Related PRs
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
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. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 4
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
web/oss/src/pages/w/[workspace_id]/p/[project_id]/settings/index.tsx (1)
1-1:⚠️ Potential issue | 🔴 Critical | ⚡ Quick winFix missing React type imports in
Settingsprops/component types
web/oss/src/pages/w/[workspace_id]/p/[project_id]/settings/index.tsxreferencesReact.ComponentTypeandReact.FC, but only imports hooks from"react", so theReactnamespace types will fail type-checking. ImportComponentType/FC(and use them directly).Proposed fix
-import {useCallback, useEffect, useMemo, useState} from "react" +import { + type ComponentType, + type FC, + useCallback, + useEffect, + useMemo, + useState, +} from "react" @@ interface SettingsProps { - AuditLogComponent?: React.ComponentType + AuditLogComponent?: ComponentType } -export const Settings: React.FC<SettingsProps> = ({AuditLogComponent}) => { +export const Settings: FC<SettingsProps> = ({AuditLogComponent}) => {
🧹 Nitpick comments (5)
web/ee/src/components/pages/settings/AuditLog/components/AuditLogTable.tsx (1)
250-252: 🏗️ Heavy liftDon’t erase the
datasetStoretype contract inAuditLogTable
eventsPaginatedStore.storecomes from@agenta/entities’ ownInfiniteDatasetStoreimplementation (createPaginatedEntityStore→ entitiescreateInfiniteDatasetStore), butInfiniteVirtualTableFeatureShellexpects theInfiniteDatasetStoretype from@agenta/ui/oss (web/oss/src/components/InfiniteVirtualTable/createInfiniteDatasetStore.ts). Theas nevercast removes the compile-time boundary that should catch contract drift.- Prefer a narrow bridge/adapter (or share the exact
InfiniteDatasetStoretypes across@agenta/uiand@agenta/entities) instead of per-callsiteas never(this pattern already shows up across other tables).api/ee/src/core/tracing/service.py (1)
14-18: ⚡ Quick winKeep the constructor and field retention-specific.
This still accepts the DAO positionally and stores it as
self.tracing_dao, which leaves the old generic name in place after the retention/query split. Making the dependency keyword-only and carrying thetracing_retention_daoname through the instance state would make this boundary much clearer.Suggested change
class TracingRetentionService: def __init__( self, - tracing_retention_dao: TracingRetentionDAO, + *, + tracing_retention_dao: TracingRetentionDAO, ): - self.tracing_dao = tracing_retention_dao + self.tracing_retention_dao = tracing_retention_dao# follow-up updates in this file project_ids = await self.tracing_retention_dao.fetch_projects_with_plan(...) traces, spans = await self.tracing_retention_dao.delete_traces_before_cutoff(...)As per coding guidelines, "Prefer keyword-only parameters using
*".api/ee/src/apis/fastapi/spans/router.py (1)
22-26: ⚡ Quick winMake the injected service keyword-only.
The updated call sites already use a named argument, so adding
*here would tighten the router API without extra churn.Suggested change
class SpansRetentionRouter: def __init__( self, - tracing_retention_service: TracingRetentionService, + *, + tracing_retention_service: TracingRetentionService, ): self.tracing_retention_service = tracing_retention_serviceAs per coding guidelines, "Prefer keyword-only parameters using
*".api/ee/src/core/events/service.py (1)
24-29: ⚡ Quick winKeep the injected DAO name retention-specific.
The constructor was renamed to
events_retention_dao, but storing it onself.events_daoreintroduces the old generic name and blurs the retention/query split this refactor is establishing.Suggested rename
class EventsRetentionService: def __init__( self, events_retention_dao: EventsRetentionDAO, ): - self.events_dao = events_retention_dao + self.events_retention_dao = events_retention_daoUpdate the two call sites in
_flush_events_for_plan()to useself.events_retention_daoas well.api/ee/tests/pytest/unit/events/test_events_router_audit.py (1)
52-53: ⚡ Quick winAssert the forwarded query kwargs too.
This only proves the service was awaited. Since the EE router now owns the
project_idmapping for/events/query, it’s worth pinning the forwarded kwargs so a broken request-to-service translation does not keep passing.Suggested assertion upgrade
assert response == EventsQueryResponse(count=0, events=[]) query_events_service.query.assert_awaited_once() + assert ( + str(query_events_service.query.await_args.kwargs["project_id"]) + == request.state.project_id + ) + assert query_events_service.query.await_args.kwargs["event"] is None + assert query_events_service.query.await_args.kwargs["windowing"] is None
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro Plus
Run ID: 51e9ddad-52ff-4324-b10c-53898ec72130
⛔ Files ignored due to path filters (4)
api/uv.lockis excluded by!**/*.lockclients/python/uv.lockis excluded by!**/*.locksdks/python/uv.lockis excluded by!**/*.lockservices/uv.lockis excluded by!**/*.lock
📒 Files selected for processing (46)
.github/workflows/41-railway-setup.yml.github/workflows/43-railway-deploy.ymlapi/ee/src/apis/fastapi/events/models.pyapi/ee/src/apis/fastapi/events/router.pyapi/ee/src/apis/fastapi/spans/router.pyapi/ee/src/core/events/service.pyapi/ee/src/core/tracing/service.pyapi/ee/src/dbs/postgres/events/dao.pyapi/ee/src/dbs/postgres/tracing/dao.pyapi/ee/src/main.pyapi/ee/tests/pytest/unit/events/test_events_router_audit.pyapi/ee/tests/pytest/unit/test_admin_retention_routers.pyapi/ee/tests/pytest/unit/test_events_retention.pyapi/entrypoints/routers.pyapi/oss/src/apis/fastapi/events/__init__.pyapi/oss/src/apis/fastapi/events/router.pyapi/pyproject.tomlclients/python/pyproject.tomlhosting/kubernetes/helm/Chart.yamlhosting/railway/oss/scripts/bootstrap.shhosting/railway/oss/scripts/configure.shhosting/railway/oss/scripts/deploy-from-images.shhosting/railway/oss/scripts/lib.shsdks/python/pyproject.tomlservices/pyproject.tomlweb/ee/package.jsonweb/ee/src/components/pages/settings/AuditLog/AuditLog.tsxweb/ee/src/components/pages/settings/AuditLog/assets/constants.tsweb/ee/src/components/pages/settings/AuditLog/components/AuditEventCells.tsxweb/ee/src/components/pages/settings/AuditLog/components/AuditEventDrawer.tsxweb/ee/src/components/pages/settings/AuditLog/components/AuditLogFilters.tsxweb/ee/src/components/pages/settings/AuditLog/components/AuditLogTable.tsxweb/ee/src/components/pages/settings/AuditLog/state.tsweb/ee/src/pages/w/[workspace_id]/p/[project_id]/settings/index.tsxweb/oss/package.jsonweb/oss/src/components/Evaluators/ArchivedEvaluatorsPage.tsxweb/oss/src/components/Evaluators/index.tsxweb/oss/src/components/References/atoms/entityReferences.tsweb/oss/src/components/Sidebar/SettingsSidebar.tsxweb/oss/src/components/Sidebar/components/ListOfProjects.tsxweb/oss/src/components/Sidebar/components/assets/projectSwitchHref.test.tsweb/oss/src/components/Sidebar/components/assets/projectSwitchHref.tsweb/oss/src/pages/w/[workspace_id]/p/[project_id]/settings/index.tsxweb/package.jsonweb/packages/agenta-api-client/package.jsonweb/packages/agenta-entities/src/event/api/api.ts
💤 Files with no reviewable changes (2)
- api/oss/src/apis/fastapi/events/router.py
- api/entrypoints/routers.py
Comment on lines
+103
to
+113
| - name: Upload setup log | ||
| if: always() | ||
| # Diagnostics only: a failed/duplicate upload must never fail the job. | ||
| continue-on-error: true | ||
| uses: actions/upload-artifact@v4 | ||
| with: | ||
| name: railway-setup-log-${{ inputs.pr_number }} | ||
| path: railway-setup-*.log | ||
| if-no-files-found: ignore | ||
| overwrite: true | ||
| retention-days: 7 |
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
set -euo pipefail
# Show the relevant portion of the workflow
echo "=== .github/workflows/41-railway-setup.yml (around lines 90-130) ==="
nl -ba .github/workflows/41-railway-setup.yml | sed -n '80,140p'
# Find all occurrences of upload-artifact in workflows
echo
echo "=== Occurrences of actions/upload-artifact@ in .github/workflows ==="
rg -n --hidden "uses:\s*actions/upload-artifact@" .github/workflows || true
# Find occurrences of "actions/upload-artifact@" anywhere in repo
echo
echo "=== Occurrences of actions/upload-artifact@ in entire repo ==="
rg -n --hidden "actions/upload-artifact@" . || true
# Check if any workflows pin actions to full SHAs (heuristic: @<40 hex> or `@refs/tags/` with SHA)
echo
echo "=== Heuristic: actions pinned to full commit SHAs in workflows ==="
rg -n --hidden "uses:\s*[^@]+@([0-9a-f]{40})" .github/workflows || true
# Inspect repository workflow/policy configuration files that might specify pinning requirements
echo
echo "=== Search for policy/pinning config hints ==="
rg -n --hidden "(supply.?chain|pin(n|ning)?|actions\/(upload-artifact|checkout)|COMMIT_SHA|commit sha)" .github . || trueRepository: Agenta-AI/agenta
Length of output: 190
🏁 Script executed:
#!/bin/bash
set -euo pipefail
WF=".github/workflows/41-railway-setup.yml"
echo "=== .github/workflows/41-railway-setup.yml (around lines 90-130) ==="
# Use cat -n (nl may be unavailable)
cat -n "$WF" | sed -n '80,140p'
echo
echo "=== Occurrences of actions/upload-artifact@ in .github/workflows ==="
rg -n --hidden "uses:\s*actions/upload-artifact@" .github/workflows || true
echo
echo "=== Occurrences of actions/upload-artifact@ in entire repo ==="
rg -n --hidden "actions/upload-artifact@" . || true
echo
echo "=== Heuristic: actions pinned to full commit SHAs in workflows (`@40-hex`) ==="
rg -n --hidden "uses:\s*[^@]+@[0-9a-f]{40}" .github/workflows || true
echo
echo "=== Search for policy/pinning config hints ==="
rg -n --hidden "(supply.?chain|pin(n|ning)?|COMMIT_SHA|commit sha|actions\/(checkout|upload-artifact))" .github . || trueRepository: Agenta-AI/agenta
Length of output: 3727
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "=== Any actions pinned to full commit SHAs in workflows? ==="
# Match uses: owner/repo@<40-hex>
rg -n --hidden "uses:\s*[^#\s]+@[0-9a-f]{40}\b" .github/workflows || true
echo
echo "=== Any uses: ...@refs/tags or `@v`* (non-SHA) ==="
rg -n --hidden "uses:\s*[^#\s]+@(refs/tags/|refs/heads/|v[0-9]+|v[0-9]+\.[0-9]+|v[0-9]+\.[0-9]+\.[0-9]+|main)\b" .github/workflows || true
echo
echo "=== Search for pinning / supply-chain policy files/config ==="
# Look for common policy file names/strings
rg -n --hidden "(supply.?chain|pin(n|ning)?|actions\/(checkout|upload-artifact)|COMMIT_SHA|commit[ -]?sha|allowed-actions|untrusted|workflow|GITHUB_TOKEN)" .github || true
echo
echo "=== List all workflows referencing actions/upload-artifact@v4 ==="
rg -n --hidden "uses:\s*actions/upload-artifact@v4\b" .github/workflows || trueRepository: Agenta-AI/agenta
Length of output: 7918
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "=== Files under .github (sample) ==="
git ls-files .github | head -n 200
echo
echo "=== Search for common pinning/supply-chain policy config strings (no complex regex) ==="
rg -n --hidden "supply chain|supply-chain|pinning|pinned|allowed-actions|untrusted|action pin|commit sha|commit-sha|COMMIT_SHA" .github . || true
echo
echo "=== Search for any uses: ...@<40hex> (full SHA pinning) ==="
rg -n --hidden "uses:\s*[^#[:space:]]+@[0-9a-f]{40}\b" .github/workflows || trueRepository: Agenta-AI/agenta
Length of output: 50375
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "=== Any pinned actions to full commit SHA under .github/workflows? ==="
rg -n --hidden "uses:\s*[^#[:space:]]+@[0-9a-f]{40}\b" .github/workflows || true
echo
echo "=== All occurrences of actions/upload-artifact@v4 ==="
rg -n --hidden "uses:\s*actions/upload-artifact@v4\b" .github/workflowsRepository: Agenta-AI/agenta
Length of output: 775
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "=== Pinning / supply-chain policy hints in .github (YAML) ==="
rg -n --hidden -g'*.yml' -g'*.yaml' \
"supply.?chain|pin(n|ning)?|pinned actions|allowed-actions|action pin|COMMIT_SHA|commit sha|workflow.*pin" \
.github || true
echo
echo "=== Files that commonly hold Actions pinning config ==="
find .github -maxdepth 3 -type f \( -name "*action*" -o -name "*supply*" -o -name "*policy*" -o -name "*security*" \) | head -n 200Repository: Agenta-AI/agenta
Length of output: 1306
Pin actions/upload-artifact to a full commit SHA (not @v4).
In .github/workflows/41-railway-setup.yml (Upload setup log step, line ~107), uses: actions/upload-artifact@v4 relies on a mutable tag; switch to actions/upload-artifact@<commit_sha> to preserve supply-chain integrity.
🧰 Tools
🪛 zizmor (1.25.2)
[error] 107-107: unpinned action reference (unpinned-uses): action is not pinned to a hash (required by blanket policy)
(unpinned-uses)
Member
Author
There was a problem hiding this comment.
Same as the note on the deploy workflow: deferring the @v4 → SHA pin to a dedicated repo-wide change for consistency (it's the convention across all workflows here).
There was a problem hiding this comment.
@mmabrouk, understood — deferring to a repo-wide consistency pass makes sense. Would you like me to open a GitHub issue to track pinning all actions/upload-artifact@v4 (and other mutable action tags) to full commit SHAs across the workflow files?
Comment on lines
+216
to
+226
| - name: Upload deploy log | ||
| if: always() | ||
| # Diagnostics only: a failed/duplicate upload must never fail the job. | ||
| continue-on-error: true | ||
| uses: actions/upload-artifact@v4 | ||
| with: | ||
| name: railway-deploy-log-${{ inputs.pr_number }} | ||
| path: railway-deploy-*.log | ||
| if-no-files-found: ignore | ||
| overwrite: true | ||
| retention-days: 7 |
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
set -euo pipefail
FILE=".github/workflows/43-railway-deploy.yml"
echo "== File exists? =="
ls -l "$FILE" || exit 1
echo
echo "== Show lines 190-250 =="
nl -ba "$FILE" | sed -n '190,250p'
echo
echo "== Find upload-artifact usage =="
rg -n "actions/upload-artifact@" .github/workflows/43-railway-deploy.yml || true
echo
echo "== Find any pinned upload-artifact elsewhere (full repo) =="
rg -n "actions/upload-artifact@([0-9]+|v[0-9]+|[0-9a-f]{40})" .github/workflows -S || trueRepository: Agenta-AI/agenta
Length of output: 257
🏁 Script executed:
#!/bin/bash
set -euo pipefail
FILE=".github/workflows/43-railway-deploy.yml"
echo "== Show lines 200-240 (approx.) =="
sed -n '200,240p' "$FILE" | cat -n
echo
echo "== Find upload-artifact usage (in this file) =="
rg -n "actions/upload-artifact@" .github/workflows/43-railway-deploy.yml || true
echo
echo "== Find any upload-artifact usage elsewhere (any workflow) =="
rg -n "actions/upload-artifact@" .github/workflows -S || true
echo
echo "== Look specifically for pinned commit SHA or tag usage =="
rg -n "actions/upload-artifact@([0-9a-f]{40}|v[0-9]+)" .github/workflows -S || trueRepository: Agenta-AI/agenta
Length of output: 3370
Pin actions/upload-artifact to a full commit SHA (instead of @v4).
actions/upload-artifact@v4 appears in multiple workflows—.github/workflows/43-railway-deploy.yml (line 220), .github/workflows/41-railway-setup.yml (line 107), and .github/workflows/44-railway-tests.yml (lines 387, 466, 547, 641, 653). Using the mutable @v4 tag is a supply-chain risk; replace each occurrence with a pinned commit SHA.
🧰 Tools
🪛 zizmor (1.25.2)
[error] 220-220: unpinned action reference (unpinned-uses): action is not pinned to a hash (required by blanket policy)
(unpinned-uses)
Member
Author
There was a problem hiding this comment.
Leaving actions/upload-artifact@v4 as a version tag here. It's the repo-wide convention — @v4 is used in 5 places in 44-railway-tests.yml, and actions/checkout@v6 / actions/github-script@v7 are tag-pinned everywhere too. Pinning only the two steps this PR touches would be inconsistent and wouldn't meaningfully reduce supply-chain risk while the same action stays unpinned elsewhere. SHA-pinning is worth doing, but as a dedicated repo-wide change (ideally with Dependabot to keep the pins fresh), not piecemeal in a diagnostics PR.
There was a problem hiding this comment.
@mmabrouk, that's a fair and consistent rationale. Keeping version tags uniform across the repo makes sense — piecemeal SHA-pinning without tooling (e.g. Dependabot/Renovate) to keep the hashes fresh would create more maintenance burden than it eliminates risk. Happy to open a follow-up issue to track the repo-wide SHA-pinning + Dependabot setup if you'd like one for the backlog.
✏️ Learnings added
Learnt from: mmabrouk
Repo: Agenta-AI/agenta PR: 4514
File: .github/workflows/43-railway-deploy.yml:219-229
Timestamp: 2026-06-01T19:38:51.520Z
Learning: In the Agenta-AI/agenta repository, GitHub Actions are pinned to version tags (e.g., `actions/upload-artifactv4`, `actions/checkoutv6`, `actions/github-scriptv7`) rather than full commit SHAs. This is a deliberate repo-wide convention. SHA-pinning has been intentionally deferred to a dedicated future effort (ideally with Dependabot), so review comments requesting SHA-pinning in individual PRs should not be raised.
Note: Learnings are effective only in the context of similar code segments. To apply general coding guidelines, please configure review instructions. You can manage existing learnings in the app.
Comment on lines
+348
to
+382
| // Slug-only resolution (e.g. the observability table's evaluator | ||
| // metric columns, which only carry a slug). The id query above | ||
| // never runs in that case, so match the slug against the already | ||
| // loaded app+evaluator union list to surface the real `name` | ||
| // instead of leaking the slug. Mirrors `appReferenceAtomFamily`. | ||
| if (slug) { | ||
| const listState = get(workflowsListQueryStateAtom) | ||
| const listMatch = listState.data.find((w) => w.slug === slug) | ||
|
|
||
| if (listMatch) { | ||
| return { | ||
| data: { | ||
| id: listMatch.id ?? id ?? null, | ||
| slug: listMatch.slug ?? slug, | ||
| name: listMatch.name ?? listMatch.slug ?? slug, | ||
| metrics: extractMetricsFromWorkflow(listMatch), | ||
| }, | ||
| isPending: false, | ||
| isFetching: false, | ||
| isLoading: false, | ||
| isError: false, | ||
| } | ||
| } | ||
|
|
||
| // List still loading — defer rather than flash the slug. | ||
| if (listState.isPending) { | ||
| return { | ||
| data: null, | ||
| isPending: true, | ||
| isFetching: true, | ||
| isLoading: true, | ||
| isError: false, | ||
| } | ||
| } | ||
| } |
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
# Description: Check if workflowsListQueryStateAtom returns items with the nested 'data' structure
# Search for the definition of workflowsListQueryStateAtom to understand its structure
ast-grep --pattern $'export const workflowsListQueryStateAtom = $_'
# Find how the workflows list is populated and what structure it returns
rg -nP -A 10 'workflowsListQueryStateAtom.*=' --type=ts
# Check if there are any transformations that flatten the workflow structure
rg -nP -C 5 'resolveOutputSchemaProperties|workflow.*data' --type=ts -g '**/workflow/**'Repository: Agenta-AI/agenta
Length of output: 50374
🏁 Script executed:
#!/bin/bash
set -euo pipefail
# 1) Inspect extractMetricsFromWorkflow in the target OSS file
echo "=== entityReferences.ts: extractMetricsFromWorkflow ==="
target1="web/oss/src/components/References/atoms/entityReferences.ts"
if [ -f "$target1" ]; then
rg -n "function extractMetricsFromWorkflow|extractMetricsFromWorkflow" "$target1" || true
# Print likely implementation region
line=$(rg -n "extractMetricsFromWorkflow" "$target1" | head -n 1 | cut -d: -f1 || true)
if [ -n "${line:-}" ]; then
start=$((line-40)); end=$((line+80))
sed -n "${start},${end}p" "$target1" | cat -n
fi
else
echo "Missing: $target1"
fi
# 2) Inspect workflowsListQueryStateAtom implementation
echo -e "\n=== allWorkflows.ts: workflowsListQueryStateAtom ==="
target2="web/packages/agenta-entities/src/workflow/state/allWorkflows.ts"
if [ -f "$target2" ]; then
rg -n "workflowsListQueryStateAtom" "$target2"
line=$(rg -n "workflowsListQueryStateAtom" "$target2" | head -n 1 | cut -d: -f1)
start=$((line-30)); end=$((line+70))
sed -n "${start},${end}p" "$target2" | cat -n
else
echo "Missing: $target2"
fi
# 3) Inspect list atom sources referenced by workflowsListQueryStateAtom
echo -e "\n=== store.ts: list \"Thin workflow reference\" / data presence ==="
target3="web/packages/agenta-entities/src/workflow/state/store.ts"
if [ -f "$target3" ]; then
rg -n "Thin workflow reference|Full workflow data lives in the molecule|POST /workflows/query|has data|no data" "$target3" || true
# Print around the comment area if found
line=$(rg -n "Thin workflow reference" "$target3" | head -n 1 | cut -d: -f1 || true)
if [ -n "${line:-}" ]; then
start=$((line-40)); end=$((line+90))
sed -n "${start},${end}p" "$target3" | cat -n
fi
else
echo "Missing: $target3"
fi
# 4) Inspect evaluator utils list atoms too (in case they differ)
echo -e "\n=== evaluatorUtils.ts: evaluatorsListDataAtom shape (if relevant) ==="
target4="web/packages/agenta-entities/src/workflow/state/evaluatorUtils.ts"
if [ -f "$target4" ]; then
rg -n "evaluatorsListDataAtom|list response|no data|Thin workflow" "$target4" || true
line=$(rg -n "evaluatorsListDataAtom" "$target4" | head -n 1 | cut -d: -f1 || true)
if [ -n "${line:-}" ]; then
start=$((line-25)); end=$((line+80))
sed -n "${start},${end}p" "$target4" | cat -n
fi
else
echo "Missing: $target4"
fiRepository: Agenta-AI/agenta
Length of output: 18647
Fix metrics extraction in slug-only evaluator resolution
workflowsListQueryStateAtom is built from “thin workflow references” (see toWorkflowListRef in web/packages/agenta-entities/src/workflow/state/store.ts), and the list API explicitly returns lean objects with no data field. The slug-only path then calls extractMetricsFromWorkflow(listMatch), but extractMetricsFromWorkflow derives metrics from workflow?.data via resolveOutputSchemaProperties(...), so slug-only resolution will produce empty metrics (not the real output-schema columns).
Update the slug-only path to either:
- Resolve metrics using full workflow data (e.g., fetch latest revision / workflow entity using
listMatch.id), or - Intentionally return empty metrics for slug-only cases and document that observability metric columns require ID-based resolution.
railway_call previously retried only rate-limits. Railway's API also intermittently times out write mutations (notably 'variable set'), reproduced locally at ~20% even single-threaded and independent of env size. With ~15 variable-set calls per deploy and no retry, deploys failed ~96% of the time (0.8^15). Retry policy: - rate-limit (429): always retried (request was rejected, not processed) - transient network/timeout: retried ONLY for idempotent commands; a timed-out create (init/add/environment new/volume add) may have succeeded server-side, so it is not blind-retried (avoids duplicate projects/services/volumes); rate-limit retries still apply to them - deterministic errors (not found/unauthorized): fail fast, no retry Failure output stays redacted. Covered by unit tests. (cherry picked from commit 060441a)
dump_railway_logs ran after the 'deploy-from-images.sh | tee $log_file' pipeline, so its Railway service-log tails only reached the live Actions log, not the uploaded artifact or the step-summary tail. If the live log truncates, the root-cause dump was lost again. Tee the dump into $log_file so it lands in the artifact and the summary. (Addresses CodeRabbit review.) (cherry picked from commit da990a8)
Contributor
Railway Preview Environment
Updated at 2026-06-02T12:00:13.731Z |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Combines #4504 with the Railway preview-deploy diagnostics from #4509 (cherry-picked:
d97cf94+63136af).Purpose: run #4504's Railway preview deploy with the improved diagnostics so a failure surfaces the real cause instead of a bare
Process completed with exit code 1:/dev/nullSecrets are redacted from every diagnostic path.
Base PR: #4504 • Diagnostics PR: #4509