ci: reduce PR feedback loop with targeted caching#1294
Merged
jeremyeder merged 8 commits intoambient-code:mainfrom Apr 17, 2026
Merged
ci: reduce PR feedback loop with targeted caching#1294jeremyeder merged 8 commits intoambient-code:mainfrom
jeremyeder merged 8 commits intoambient-code:mainfrom
Conversation
Contributor
📝 WalkthroughWalkthroughPins multiple GitHub Action references to commit SHAs; refactors E2E image preparation into per-image conditional BuildKit build vs pull-and-retag steps with cache scopes and a “show built images” step; adds Changes
sequenceDiagram
actor Workflow
participant Detect as "detect-changes"
participant Builder as "docker/build-push-action"
participant Registry as "Container Registry"
participant Cache as "GHA cache (layers)"
participant KindCache as "actions/cache (kind)"
participant Runner as "Job runner"
Workflow->>Detect: run per-image detection
Detect-->>Workflow: outputs (changed / unchanged)
alt changed
Workflow->>Builder: build image (cache-from/cache-to)
Builder->>Cache: pull/push layer caches
Builder->>Registry: push :e2e-test
else unchanged
Workflow->>Registry: docker pull :latest
Registry-->>Workflow: image
Workflow->>Registry: docker tag -> :e2e-test
end
Workflow->>Runner: show built images
%% kind binary flow
Workflow->>KindCache: restore key(KIND_VERSION)
alt cache hit
KindCache-->>Workflow: kind present
else cache miss
Workflow->>Registry: download kind (using KIND_VERSION)
Workflow->>KindCache: save kind to cache
end
🚥 Pre-merge checks | ✅ 6✅ Passed checks (6 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
✨ Simplify code
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
.github/workflows/lint.yml (1)
144-149:⚠️ Potential issue | 🟠 MajorBackend lint skips production-only files with
--build-tags=test.
components/backend/handlers/k8s_clients_for_request_prod.gohas//go:build !testand won't be linted by the current invocation. Add a second pass without build-tag restriction to catch production-only code.Proposed fix
- name: Run golangci-lint (all build tags) uses: golangci/golangci-lint-action@v9 with: version: latest working-directory: components/backend args: --timeout=5m --build-tags=test + - name: Run golangci-lint (default build tags) + uses: golangci/golangci-lint-action@v9 + with: + version: latest + working-directory: components/backend + args: --timeout=5m🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In @.github/workflows/lint.yml around lines 144 - 149, CI currently runs golangci-lint only with --build-tags=test, which skips production-only files like components/backend/handlers/k8s_clients_for_request_prod.go (//go:build !test); add a second golangci-lint step that also runs in the same working-directory (components/backend) but without --build-tags (or with an explicit empty/omitted build-tags arg) so production-only files are linted too, keeping the existing step for test-tagged checks; target the existing action block name "Run golangci-lint (all build tags)" or add a new step named e.g. "Run golangci-lint (no build tags)" to make this explicit.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In @.github/workflows/e2e.yml:
- Around line 12-17: Add an explicit permissions block at the top-level of this
workflow (near the existing env/KIND_VERSION and concurrency keys) that scopes
GITHUB_TOKEN to only the least-privilege scopes required for the job (do NOT
rely on default write); for example set only the specific permissions you need
such as contents: read, actions: read, id-token: write, or packages: write as
appropriate for your cache/action usage, and remove any broader write defaults.
Also pin every third-party action used in the workflow to an exact commit SHA
(instead of floating tags) so the run is hermetic and auditable.
- Around line 95-96: Replace all mutable action tags with immutable commit SHAs
for each action usage (e.g., replace docker/build-push-action@v7 and the other
`@v6/`@v4/@v3 occurrences with their corresponding full commit SHAs) so every
"uses:" entry is pinned; update the 12 action instances referenced (including
docker/build-push-action and the other actions in the file) to their exact SHA
values. Also add a top-level permissions: block to the workflow that scopes the
GITHUB_TOKEN to the minimal required permissions for this workflow (declare only
the specific permission keys needed, e.g., read/write for specific resources
used), ensuring the token is not granted broad defaults. Ensure changes touch
the workflow root (top-level) and every "uses:" line that currently has a
version tag.
---
Outside diff comments:
In @.github/workflows/lint.yml:
- Around line 144-149: CI currently runs golangci-lint only with
--build-tags=test, which skips production-only files like
components/backend/handlers/k8s_clients_for_request_prod.go (//go:build !test);
add a second golangci-lint step that also runs in the same working-directory
(components/backend) but without --build-tags (or with an explicit empty/omitted
build-tags arg) so production-only files are linted too, keeping the existing
step for test-tagged checks; target the existing action block name "Run
golangci-lint (all build tags)" or add a new step named e.g. "Run golangci-lint
(no build tags)" to make this explicit.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: a3e90040-f7a9-491d-bf5b-3d9c35e7ab58
📒 Files selected for processing (5)
.github/workflows/e2e.yml.github/workflows/lint.yml.github/workflows/unit-tests.ymldocs/superpowers/plans/2026-04-11-ci-improvements.mddocs/superpowers/specs/2026-04-11-ci-improvements-design.md
Targeted caching wins across all PR workflows plus E2E image reuse via shared GHA BuildKit cache scopes. Goal: cut PR wall-clock from ~10.4m P50 to ~5-7m. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
4 tasks: E2E Docker layer caching (shared scopes from components-build), kind binary caching, golangci-lint consolidation, junit2html via pipx. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The test build tag is a superset: linting with --build-tags=test covers all production files plus test-tagged files. Reduces CI runtime by eliminating redundant pass. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace monolithic shell script with individual docker/build-push-action
steps. Each component build now reads GHA cache from components-build-deploy
workflow (scope: {component}-amd64) and writes to e2e-specific scope
(scope: e2e-{component}) for future runs.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add workflow-level KIND_VERSION env var and use actions/cache@v4 to store the kind binary in ~/k8s-tools/kind. Cache key includes OS and version for invalidation on upgrades. Avoids redundant downloads on every workflow run. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Pin all 13 action references in e2e.yml to commit SHAs - Pin all action references in lint.yml and unit-tests.yml to commit SHAs - Add explicit permissions block (contents: read, actions: write) to e2e.yml Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
jeremyeder
force-pushed
the
feature/ci-improvements
branch
from
bcdfbc7 to
7c9a714
Compare
✅ Deploy Preview for cheerful-kitten-f556a0 canceled.
|
v3.12.1 release artifact is returning truncated gzip, breaking CI. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In @.github/workflows/docs-lint.yml:
- Around line 16-19: Add an explicit least-privilege permissions block for the
workflow (or specific job) so GITHUB_TOKEN is scoped; update the workflow
containing uses: actions/checkout and uses: actions/setup-node to include a
permissions: block with at minimum contents: read at the top-level of the
workflow or inside the job that runs these actions to ensure the token cannot
write by default.
- Line 27: The workflow step that runs "curl -sfL
https://github.com/errata-ai/vale/releases/download/v3.14.1/vale_3.14.1_Linux_64-bit.tar.gz
| tar xz -C /usr/local/bin vale" lacks integrity checks; change it to first
download both the archive and the corresponding checksum file
"vale_3.14.1_checksums.txt", compute/verify the SHA-256 checksum (e.g., with
sha256sum -c or by comparing sha256sum output) against the expected entry in the
checksum file, and only call tar xz -C /usr/local/bin vale after the checksum
verification succeeds, failing the job if the checksum does not match.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro Plus
Run ID: 1398b3cf-b8e9-40c6-926e-658687d8a2a6
📒 Files selected for processing (1)
.github/workflows/docs-lint.yml
Comment on lines
+16
to
+19
| uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2 | ||
|
|
||
| - name: Setup Node.js | ||
| uses: actions/setup-node@v6 | ||
| uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6 |
Contributor
There was a problem hiding this comment.
Add explicit least-privilege permissions for this workflow.
This workflow does not scope GITHUB_TOKEN permissions explicitly. Add a minimal permissions block (contents: read) at workflow or job level.
Proposed minimal permission scope
name: Docs Lint
on:
pull_request:
paths:
- 'docs/**'
+
+permissions:
+ contents: read
jobs:
lint:
name: Lint Documentation🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In @.github/workflows/docs-lint.yml around lines 16 - 19, Add an explicit
least-privilege permissions block for the workflow (or specific job) so
GITHUB_TOKEN is scoped; update the workflow containing uses: actions/checkout
and uses: actions/setup-node to include a permissions: block with at minimum
contents: read at the top-level of the workflow or inside the job that runs
these actions to ensure the token cannot write by default.
| - name: Install Vale | ||
| run: | | ||
| curl -sfL https://github.com/errata-ai/vale/releases/download/v3.12.1/vale_3.12.1_Linux_64-bit.tar.gz | tar xz -C /usr/local/bin vale | ||
| curl -sfL https://github.com/errata-ai/vale/releases/download/v3.14.1/vale_3.14.1_Linux_64-bit.tar.gz | tar xz -C /usr/local/bin vale |
Contributor
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
fd -type f -name "docs-lint.yml" .github/workflows/Repository: ambient-code/platform
Length of output: 235
🏁 Script executed:
cat -n .github/workflows/docs-lint.yml | sed -n '20,35p'Repository: ambient-code/platform
Length of output: 640
🌐 Web query:
Vale errata-ai releases checksum SHA verification GitHub
💡 Result:
Vale (from Errata AI) provides SHA-256 checksums for verifying downloaded release binaries on GitHub at https://github.com/errata-ai/vale/releases. Each release includes a vale_{version}_checksums.txt file listing SHA-256 hashes for all platform-specific archives (e.g., Linux, macOS, Windows tar.gz/zip files). To verify: 1. Download the desired release assets from a tag page, e.g., https://github.com/errata-ai/vale/releases/tag/v3.13.1 2. Download vale_3.13.1_checksums.txt 3. Run: sha256sum -c vale_3.13.1_checksums.txt (Linux/macOS) or equivalent on Windows (e.g., certutil -hashfile). Releases are GPG-signed commits (e.g., key ID B44A6E0381723797). Official docs recommend binary downloads but do not explicitly detail checksum verification steps; it follows standard GitHub practices. No separate PGP-signed binaries or Cosign noted.
Citations:
- 1: https://github.com/errata-ai/vale/releases
- 2: https://github.com/errata-ai/vale/releases/tag/v3.9.2
- 3: https://github.com/errata-ai/vale/releases/tag/v3.13.1
- 4: https://github.com/errata-ai/vale/releases/tag/v2.16.0
- 5: https://mintlify.com/errata-ai/vale/installation
Add checksum verification for Vale binary.
curl | tar downloads and extracts without integrity validation. Vale releases include SHA-256 checksums at vale_3.14.1_checksums.txt—use them to verify the archive before extraction.
Recommended pattern
- name: Install Vale
run: |
- curl -sfL https://github.com/errata-ai/vale/releases/download/v3.14.1/vale_3.14.1_Linux_64-bit.tar.gz | tar xz -C /usr/local/bin vale
+ VALE_VERSION="3.14.1"
+ VALE_TGZ="/tmp/vale_${VALE_VERSION}_Linux_64-bit.tar.gz"
+ VALE_URL="https://github.com/errata-ai/vale/releases/download/v${VALE_VERSION}/vale_${VALE_VERSION}_Linux_64-bit.tar.gz"
+ curl -fsSL -o "${VALE_TGZ}" "${VALE_URL}"
+ curl -fsSL "https://github.com/errata-ai/vale/releases/download/v${VALE_VERSION}/vale_${VALE_VERSION}_checksums.txt" | grep "Linux_64-bit.tar.gz" | sha256sum -c -
+ tar -xzf "${VALE_TGZ}" -C /usr/local/bin vale📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| curl -sfL https://github.com/errata-ai/vale/releases/download/v3.14.1/vale_3.14.1_Linux_64-bit.tar.gz | tar xz -C /usr/local/bin vale | |
| VALE_VERSION="3.14.1" | |
| VALE_TGZ="/tmp/vale_${VALE_VERSION}_Linux_64-bit.tar.gz" | |
| VALE_URL="https://github.com/errata-ai/vale/releases/download/v${VALE_VERSION}/vale_${VALE_VERSION}_Linux_64-bit.tar.gz" | |
| curl -fsSL -o "${VALE_TGZ}" "${VALE_URL}" | |
| curl -fsSL "https://github.com/errata-ai/vale/releases/download/v${VALE_VERSION}/vale_${VALE_VERSION}_checksums.txt" | grep "Linux_64-bit.tar.gz" | sha256sum -c - | |
| tar -xzf "${VALE_TGZ}" -C /usr/local/bin vale |
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In @.github/workflows/docs-lint.yml at line 27, The workflow step that runs
"curl -sfL
https://github.com/errata-ai/vale/releases/download/v3.14.1/vale_3.14.1_Linux_64-bit.tar.gz
| tar xz -C /usr/local/bin vale" lacks integrity checks; change it to first
download both the archive and the corresponding checksum file
"vale_3.14.1_checksums.txt", compute/verify the SHA-256 checksum (e.g., with
sha256sum -c or by comparing sha256sum output) against the expected entry in the
checksum file, and only call tar xz -C /usr/local/bin vale after the checksum
verification succeeds, failing the job if the checksum does not match.
jeremyeder
added a commit
that referenced
this pull request
Apr 21, 2026
PR #1294 reintroduced context: components/runners, but PR #1260 had already changed the Dockerfile to COPY . (expecting the context to be the ambient-runner subdirectory). This caused pip3 install to fail with "Neither setup.py nor pyproject.toml found". Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
jeremyeder
added a commit
that referenced
this pull request
Apr 21, 2026
PR #1294 reintroduced context: components/runners, but PR #1260 had already changed the Dockerfile to COPY . (expecting the context to be the ambient-runner subdirectory). This caused pip3 install to fail with "Neither setup.py nor pyproject.toml found". Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
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
1 participant
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.
Summary
docker buildwithdocker/build-push-action@v7using GHA cache. Reads layers from the components-build-deploy workflow's cache scopes (frontend-amd64, etc.) so E2E gets warm layers from the last main build. Expected savings: 3-5 min.actions/cache@v4. Pin version in env var for cache key stability. ~15s saved per run.--build-tags=test(superset). ~30s saved.pip install junit2htmlwithpipx run junit2html(pre-installed on GHA runners). ~10s saved.Current PR wall-clock P50: ~10.4m (E2E bottleneck)
Expected PR wall-clock P50: ~5-7m
Test plan
:latestfor unchanged components🤖 Generated with Claude Code
Summary by CodeRabbit
Chores
Documentation