Skip to content

ci: route L4 step-name and evidence-mirror checks to optional - #63

Merged
jadoonf merged 2 commits into
mainfrom
devin/1789171000-l4-carveouts
Sep 12, 2026
Merged

jadoonf merged 2 commits into
mainfrom
devin/1789171000-l4-carveouts

Conversation

@jadoonf

@jadoonf jadoonf commented Sep 11, 2026 •

Copy link
Copy Markdown

garnet-gate: jibril_version=v2.17.0

Summary

Two L4 sub-checks move to optional per the rollout decision, plus two gate-side grep false-negatives fixed — all observed on the first tiered run (#61's combination verdict):

  • Quoted step name in the comment body (F31) — step attribution is a disclosed-unstable surface (F25); the docs call it out as such. Missing quoted name → optional. A missing destination stays core — that's the comment's actual claim. A second Garnet App comment on one head (update-in-place violation) stays core too.
  • PR-description evidence mirror (F34) — a reviewer-consumption experiment, not part of the pnpm acceptance surface. Absent-or-stale mirror → optional in both directions.
  • Rendered-body normalization — the App comment defangs domains (registry.npmjs[.]org) and drops jibril's 2. step-numbering prefix, so raw greps for registry.npmjs.org and "2. Stop the action clock" missed evidence that renders correctly. The check now undefangs the body and tries the de-numbered step name — it reads what a reader sees.

leg-tiers.yml L4 why carries the carve-out so the config stays the source of truth for where the bar sits.

Squash Commit Body

Route L4's quoted-step-name and evidence-mirror misses to optional: step
attribution is disclosed-unstable and the description mirror is a
reviewer-consumption experiment. A missing destination or a duplicate App
comment on one head still fails the leg. Normalize the rendered comment
body before grepping it — undefang `[.]` domains and drop jibril's
step-numbering prefix — so rendered evidence is not counted absent.

Checklist

  • Gate label garnet-release-testing applied; this PR dogfoods the change.
  • Gate verdict shows the carve-outs as OPT and L4's destination/step checks pass on rendered content.

View with [code]smith Autofix with [code]smith
Need help on this PR? Tag @codesmith-bot with what you need. Autofix is disabled.

Runtime evidence (Garnet)

Kernel-recorded execution record for head afc58c16d097a9c34eefb94f3b342b076a47129b, mirrored verbatim from
the sticky Garnet Runtime Review comment on this PR so reviewers that read only
the description ground in the same bytes. Facts only. Judgment stays with the
reviewer. Cite grounded findings as:

Runtime evidence (Garnet, head afc58c1): <execution chain> → <destination> (<workflow>/<job>) —

Execution record (verbatim mirror)

Execution Profiles recorded for 2 jobs, triggered by afc58c1

No changes since 470851f
recorded at the kernel by Garnet · 2026-09-11 23:57 UTC

Garnet Jibril Release Gate / reproduce ↗ · 3 destinations · unchanged
Runner.Worker
└─ bash
   └─ curl (step: "Stop the action clock")
      ├─ ○ localhost (dns resolver)
      └─ ○ registry.npmjs[.]org

systemd
└─ systemd-resolved
   └─ ○ 8.8.8.8

View this job's Execution Profile in Garnet →

TS CI / test ↗ · 8 destinations · unchanged
Runner.Worker
└─ node
   ├─ pnpm
   │  ├─ ○ nodejs[.]org
   │  └─ ○ unofficial-builds.nodejs[.]org
   ├─ ○ blacksmith.blob.core.windows[.]net
   ├─ ○ localhost (dns resolver)
   └─ ○ registry.npmjs[.]org

systemd
├─ chronyd
│  └─ ○ 169.254.169.123
├─ nginx
│  └─ nginx
│     └─ ○ 192.168.127.1
└─ systemd-resolved
   └─ ○ 8.8.8.8

View this job's Execution Profile in Garnet →


💡 How to read this
Runner.Worker          ← process on a path
└─ npm
   └─ node             ← process that acted
      └─ ○ npmjs[.]org ← observed action

follow a path downward to see what ran and what it did — each path to an observed action is an execution chain

names on the path = processes · ○ = observed action · (…) = context

+ only in the current record · − only in the previous record

Workload behaviour (Garnet)

Garnet Jibril Release Gate / reproduce · head e09e30b · 5 workload chains · 2 runner-platform chains

Workload behaviour vs ba67148: +0 −8 destinations
− github[.]com
− codeload.github[.]com
− nodejs[.]org
− node-registry.bit[.]cloud
− pnpm[.]io
− release-assets.githubusercontent[.]com
− unofficial-builds.nodejs[.]org
− npm.jsr[.]io
runner platform: +0 −19 destinations · not workload behaviour

5 workload chains · 2 destinations at head, 8 gone
destination chains Δ vs base reached by
localhost 3 19 → 3 curl, node, sudo
registry.npmjs[.]org 2 72 → 2 curl, node
codeload.github[.]com 0 gone node
github[.]com 0 gone git-remote-http, node
node-registry.bit[.]cloud 0 gone node
nodejs[.]org 0 gone node
npm.jsr[.]io 0 gone node
pnpm[.]io 0 gone node
release-assets.githubusercontent[.]com 0 gone node
unofficial-builds.nodejs[.]org 0 gone node
runner platform · 2 chains · 2 destinations at head, 19 gone · no recorded workflow step
destination chains Δ vs base reached by
8.8.8.8 1 2 → 1 systemd-resolved
localhost 1 7 → 1 sudo
140.82.112.22 0 gone nginx
140.82.113.21 0 gone nginx
140.82.114.21 0 gone nginx
169.254.169.123 0 gone chronyd
185.125.190.56 0 gone chronyd
185.125.190.57 0 gone chronyd
185.125.190.58 0 gone chronyd
192.168.127.1 0 gone blacksmithd, sudo
198.137.202.56 0 gone chronyd
23.143.196.204 0 gone chronyd
23.186.168.128 0 gone chronyd
5.161.94.12 0 gone chronyd
91.189.91.157 0 gone chronyd
api.snapcraft[.]io 0 gone snapd
cloudfront.cdn.snapcraftcontent[.]com 0 gone snapd
nodejs[.]org 0 gone pnpm
productionresultssa15.blob.core[.]windows 0 gone node
registry.npmjs[.]org 0 gone pnpm
unofficial-builds.nodejs[.]org 0 gone pnpm

recorded at the kernel by Garnet · profile view →

Summary by CodeRabbit

  • Bug Fixes
    • Improved release validation to recognize equivalent formatting for destinations and recorded step names.
    • Reduced false failures when optional supporting evidence is missing or differs.
    • Preserved enforcement of required destination and duplicate-comment checks.
    • Improved handling of empty validation records for more reliable results.

Greptile Summary

Summary

  • Updates Garnet release-gate policy handling for optional rendered-evidence checks.
  • Normalizes defanged destinations and accepts de-numbered Jibril step names during rendered-evidence verification.
  • No new actionable issues were identified.

Confidence Score: 5/5

Safe to merge; no blocking issues require changes.

There are no outstanding findings. No non-blocking concerns were identified.

Files Needing Attention: None.

Reviews (2): Last reviewed commit: "ci(garnet): normalize the rendered comme..." | Re-trigger Greptile

A missing quoted step name in the App comment and the PR-description
evidence mirror report optional: step attribution is disclosed-unstable
(F25) and the mirror is a reviewer-consumption experiment (F34). A missing
destination or a second Garnet App comment on one head stays core.
@devin-ai-integration

Copy link
Copy Markdown

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

  • Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
  • Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

  • Disable automatic comment, CI, and merge conflict monitoring

@coderabbitai

coderabbitai Bot commented Sep 11, 2026 •

Copy link
Copy Markdown

Review Change StackReview Change Stack

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info
⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Advanced

Run ID: ade73f2f-ac9c-45d7-8324-584fcbb69c3d

📥 Commits

Reviewing files that changed from the base of the PR and between bda4f64 and afc58c1.

📒 Files selected for processing (2)
  • .github/garnet-gate/leg-tiers.yml
  • .github/scripts/garnet-release-gate-verify.sh

📝 Walkthrough

Walkthrough

The release gate now normalizes L4 App-comment evidence, downgrades selected missing evidence to optional misses, updates the L4 rationale, and builds L6 shape state with empty-array-safe note iteration.

Changes

Release gate validation

Layer / File(s) Summary
L4 evidence classification
.github/garnet-gate/leg-tiers.yml, .github/scripts/garnet-release-gate-verify.sh
L4 matching normalizes rendered destinations and numbered step names. Missing destinations and duplicate App comments remain core failures. Missing step names and evidence mirrors are optional misses.
L6 shape-state assembly
.github/scripts/garnet-release-gate-verify.sh
shape_state combines shape and upstream notes with empty-array-safe iteration and trims the final separator.

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix

Suggested reviewers: zkochan

✨ Finishing Touches 💡 1
🛠️ Fix failing CI checks 💡
  • Create stacked PR
  • Commit on current branch
📝 Generate docstrings
  • Create stacked PR
  • Commit on current branch
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Commit unit tests in branch devin/1789171000-l4-carveouts

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

Comment @coderabbitai help to get the list of available commands.

@devin-ai-integration devin-ai-integration Bot added the garnet-release-testing Garnet product/release test PR; the Jibril release gate runs and reports here label Sep 11, 2026
@garnet-runtime-review

garnet-runtime-review Bot commented Sep 11, 2026 •

Copy link
Copy Markdown

Execution Profiles recorded for 2 jobs, triggered by afc58c1

No changes since 470851f
recorded at the kernel by Garnet · 2026-09-11 23:57 UTC

Garnet Jibril Release Gate / reproduce ↗ · 3 destinations · unchanged
Runner.Worker
└─ bash
   └─ curl (step: "Stop the action clock")
      ├─ ○ localhost (dns resolver)
      └─ ○ registry.npmjs[.]org

systemd
└─ systemd-resolved
   └─ ○ 8.8.8.8

View this job's Execution Profile in Garnet →

TS CI / test ↗ · 8 destinations · unchanged
Runner.Worker
└─ node
   ├─ pnpm
   │  ├─ ○ nodejs[.]org
   │  └─ ○ unofficial-builds.nodejs[.]org
   ├─ ○ blacksmith.blob.core.windows[.]net
   ├─ ○ localhost (dns resolver)
   └─ ○ registry.npmjs[.]org

systemd
├─ chronyd
│  └─ ○ 169.254.169.123
├─ nginx
│  └─ nginx
│     └─ ○ 192.168.127.1
└─ systemd-resolved
   └─ ○ 8.8.8.8

View this job's Execution Profile in Garnet →


💡 How to read this
Runner.Worker          ← process on a path
└─ npm
   └─ node             ← process that acted
      └─ ○ npmjs[.]org ← observed action

follow a path downward to see what ran and what it did — each path to an observed action is an execution chain

names on the path = processes · ○ = observed action · (…) = context

+ only in the current record · − only in the previous record

The comment defangs domains (`registry.npmjs[.]org`) and drops jibril's
step-numbering prefix, so raw-body greps for `registry.npmjs.org` and
`"2. Stop the action clock"` missed rendered-but-present evidence.
Normalize the body (`[.]` -> `.`) and try the de-numbered step name.
@github-actions

Copy link
Copy Markdown

/agentic_review

@github-actions

Copy link
Copy Markdown

@coderabbitai review

@qodo-code-review

qodo-code-review Bot commented Sep 11, 2026 •

Copy link
Copy Markdown

Code Review by Qodo

🐞 Bugs (1) 📘 Rule violations (1) 📜 Skill insights (0)

Grey Divider


Remediation recommended

1. Mirror comment repeats branch behavior 📘 Rule violation ⚙ Maintainability ⭐ New
Description
Reported optional in both directions narrates the two immediately following branches, which both
call leg_optional, without adding rationale beyond the preceding sentence. If either branch's tier
changes later, this duplicate summary can become stale and misdescribe the mirror policy.
Code

.github/scripts/garnet-release-gate-verify.sh[332]

+    # the pnpm acceptance surface (F34). Reported optional in both directions.
Relevance

●●● Strong

Redundant behavior-only comment conflicts with the repository’s active no-duplicate-comments rule.

PR-#43

ⓘ Recommendations generated based on similar findings in past PRs

Evidence
Compliance rule 2757161 prohibits comments that merely restate straightforward code behavior. The
added sentence at line 332 says both outcomes are optional, while lines 337-342 express that
directly through leg_optional calls.

Rule 2757161: Avoid redundant comments that duplicate code behavior
.github/scripts/garnet-release-gate-verify.sh[330-342]

Agent prompt
The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
The mirror comment directly repeats the optional routing expressed by the following branches and may become stale if that routing changes.

## Fix Focus Areas
- .github/scripts/garnet-release-gate-verify.sh[330-332]

## Recommended Fix
Remove `Reported optional in both directions.` while retaining the preceding rationale explaining why the evidence mirror is outside the acceptance surface.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools


2. Wrong destinations pass the comment gate 🐞 Bug ≡ Correctness
Description
The normalized-body check uses fixed-string substring matching for WORKLOAD_DESTINATION rather
than verifying a complete rendered hostname. A comment containing only a different destination such
as registry.npmjs[.]org.evil.example is normalized to a matching substring, allowing the core L4
check and overall gate verdict to pass without displaying the expected destination.
Code

.github/scripts/garnet-release-gate-verify.sh[318]

+    grep -qF "$WORKLOAD_DESTINATION" <<<"$body_norm" && has_dest=true
Relevance

●●● Strong

Accepted precedent prioritizes preventing gate PASS when required workload evidence is unverified;
this is a similarly exploitable false-positive.

PR-#43

ⓘ Recommendations generated based on similar findings in past PRs

Evidence
The profile path selects workload steps only when a remote name exactly equals the expected
destination, while the changed rendered-comment path reduces that stronger requirement to substring
containment. L4 is configured as core, and only accumulated core failures change the final verdict,
so this false match can suppress the failure the check is intended to produce.

.github/scripts/garnet-release-gate-verify.sh[191-195]
.github/scripts/garnet-release-gate-verify.sh[312-324]
.github/garnet-gate/leg-tiers.yml[28-33]
.github/scripts/garnet-release-gate-verify.sh[755-757]

Agent prompt
The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
The normalized comment-body check accepts the expected destination as a substring, so a longer unrelated hostname can satisfy the core rendered-evidence check.

## Fix Focus Areas
- .github/scripts/garnet-release-gate-verify.sh[312-318]

## Recommended Fix
After normalizing the defanged form, match the destination with hostname boundaries or parse the rendered destination tokens and compare them exactly to `WORKLOAD_DESTINATION`. Add coverage showing that `registry.npmjs[.]org` passes while prefixed or suffixed hostnames do not.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools


Grey Divider

Context sources
✅ Compliance rules (platform): 36 rules
✅ REVIEW.md
Review mode: ⚖️ Balanced: This changes runtime CI gate logic and security-sensitive evidence validation, including destination and head-bound comment checks; the localized diff has a known pending correctness concern but does not warrant redundant extended passes.

Grey Divider

Tip of the day
💡 Did you know, you can hide the parts of a finding you never read, like the evidence or the agent prompt

More tips ↗ | Customize Qodo ↗ | Qodo docs ↗

Grey Divider

Previous reviews

Review updated until commit afc58c1 ⚖️ Balanced

Results up to commit afc58c1 ⚖️ Balanced


🐞 Bugs (1) 📘 Rule violations (0) 📎 Requirement gaps (0) 🎨 UX issues (0) 🔗 Cross-repo conflicts (0) 📜 Skill insights (0)


Remediation recommended
1. Wrong destinations pass the comment gate 🐞 Bug ≡ Correctness
Description
The normalized-body check uses fixed-string substring matching for WORKLOAD_DESTINATION rather
than verifying a complete rendered hostname. A comment containing only a different destination such
as registry.npmjs[.]org.evil.example is normalized to a matching substring, allowing the core L4
check and overall gate verdict to pass without displaying the expected destination.
Code

.github/scripts/garnet-release-gate-verify.sh[318]

+    grep -qF "$WORKLOAD_DESTINATION" <<<"$body_norm" && has_dest=true
Relevance

●●● Strong

Accepted precedent prioritizes preventing gate PASS when required workload evidence is unverified;
this is a similarly exploitable false-positive.

PR-#43

ⓘ Recommendations generated based on similar findings in past PRs

Evidence
The profile path selects workload steps only when a remote name exactly equals the expected
destination, while the changed rendered-comment path reduces that stronger requirement to substring
containment. L4 is configured as core, and only accumulated core failures change the final verdict,
so this false match can suppress the failure the check is intended to produce.

.github/scripts/garnet-release-gate-verify.sh[191-195]
.github/scripts/garnet-release-gate-verify.sh[312-324]
.github/garnet-gate/leg-tiers.yml[28-33]
.github/scripts/garnet-release-gate-verify.sh[755-757]

Agent prompt
The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
The normalized comment-body check accepts the expected destination as a substring, so a longer unrelated hostname can satisfy the core rendered-evidence check.

## Fix Focus Areas
- .github/scripts/garnet-release-gate-verify.sh[312-318]

## Recommended Fix
After normalizing the defanged form, match the destination with hostname boundaries or parse the rendered destination tokens and compare them exactly to `WORKLOAD_DESTINATION`. Add coverage showing that `registry.npmjs[.]org` passes while prefixed or suffixed hostnames do not.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools


Grey Divider

Qodo Logo

@github-actions

github-actions Bot commented Sep 11, 2026 •

Copy link
Copy Markdown

Jibril release gate: PASS

v2.17.0 on garnet-org/action@9696f3c cleared every core leg of the pnpm acceptance bar · 3 optional leg(s) unmet (accepted limitations and follow-ons, below).

Core legs gate the pnpm upgrade verdict (capture → comment → permalink, rendered-comment correctness, security/trust posture, rollout-model compliance). OPT legs are disclosed limitations and follow-ons, or legs only pnpm's own repin can exercise — visible on every run, never blocking. Tiers live in .github/garnet-gate/leg-tiers.yml.

leg what it found
PASS L1 startup PASS — sensor active, status ok, github steps resolution ok →
PASS L2 profile uploaded after 0s · 01a092e7-c005-7dc2-a6bd-7a894edf46b3 · 3 destinations →
OPT L3 workload · optional destination present, but attributed to steps ["2. Stop the action clock"] instead of "Workload … →
PASS L4 App comment final after 1s · names "2. Stop the action clock" and registry.npmjs.org · mirrored in the desc… →
PASS L5 permalink html 200 · api 200 →
OPT L6 posture upstream test: sensor version pinned — action default at 3d47f4a resolves to latest
PASS L7 pnpm's own run · optional profile 01a092e1-3c0d-74dd-8127-3dcab856eafb from TS CI run 34659332540 (success)
PASS L8 scanners 0 regular, 0 pedantic finding(s) over 2 workflow file(s); 0 without a written class · actionlin…
PASS L9 timing sensor start 14s (budget 30s) · workload 1s with sensor vs 0s without, overhead 1s (budget 120s…
PASS L10 coverage 1 of 6 CI cells instrumented: ubuntu/node 24 · not instrumented: ubuntu/node 22, ubuntu/node 26…
PASS L11 upstream drift pnpm/pnpm at 7737634 · 0 difference(s), 2 declared deliberate →
PASS L12 warning noise 0 warning(s), 0 error(s) from the action step; 0 are the disclosed fail-open text (F2) →
OPT L13 credential-less (simulation) · optional outcome success · 0 warning(s), 0 error(s), 0 info line(s) naming the missing token →
PASS L14 release and tag · optional release.yml verify-ts-release on macos-latest: profile no (macos runner: the sensor is Linu…

Optional legs unmet (never block the verdict)

  • L3 workload: destination present, but attributed to steps ["2. Stop the action clock"] instead of "Workload with egress" — F25 Jibril step-numbering skew, garnet-org/jibril parseStepsList: the egress lands on "Stop the action clock", the run step one place earlier in the file
  • L6 posture (upstream): upstream test: sensor version pinned — action default at 3d47f4a resolves to latest;
  • L13 credential-less: the action skipped silently; a credential-less run has to say why (F28)

Disclosed, not failing

  • the sensor is Linux x86_64 eBPF: macOS and Windows cells record nothing, and no change to this rollout makes them record
  • fork pull requests and Dependabot-actor runs get no secret, so they produce no server-side evidence; L13 is a simulation of that path, not evidence for it
  • the action fails open: a sensor that never starts leaves a green job (F2), which is why the gate reads the status file and the profile rather than the job conclusion
  • no_bad_egress_domain has been attention-only in every pnpm profile and has never had a deny rule to fail against: it is not a proof of anything until pnpm has deny rules in the control plane
  • L13 is a simulation on this repository's own event with a read-only token, not a Dependabot run: it shows what the action does without a token, and no server-side evidence exists for Dependabot or fork pull requests (F11, F12)
  • release.yml runs a Garnet step in verify-ts-release on macos-latest and records nothing: macos runner: the sensor is Linux-only eBPF (F13)
  • release.yml runs a Garnet step in release on macos-latest and records nothing: macos runner: the sensor is Linux-only eBPF (F13)

Record

field value
jibril requested v2.17.0 · resolved v2.17.0
action garnet-org/action@9696f3cae144 · empty jibril_version there resolves to v2.16.0
run 34659332484 attempt 1 · blacksmith-01m29e3vpkekbew6y1hp6w8a98-8vcpu · pull_request
pnpm at main 7737634 · action 3d47f4a9004f → latest
workflow ref seen by sensor garnet-labs/pnpm/.github/workflows/garnet-jibril-release-gate.yml@refs/pull/63/merge
base → head bda4f64 → afc58c1
merge sha at run e09e30b · now e09e30b · profile e09e30b
comparison immediate parent → head (previous 470851f)
synthetic merge yes · sensor recorded merge commit e09e30b, PR head is afc58c1
permalink https://app.garnet.ai/public/runs/34659332484?profile=01a092e7-c005-7dc2-a6bd-7a894edf46b3

Legs map to past failures in .github/garnet-gate/FAILURE_LEDGER.md. Full record in the run's garnet-release-gate artifact.

@jadoonf
jadoonf marked this pull request as ready for review September 12, 2026 00:08
@jadoonf
jadoonf merged commit 7768468 into main Sep 12, 2026
34 of 37 checks passed

@devin-ai-integration devin-ai-integration Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

✅ Devin Review: No Issues Found

Devin Review analyzed this PR and found no bugs or issues to report.

Devin Review

@qodo-code-review

Copy link
Copy Markdown

PR Summary by Qodo

Route L4 carve-outs to optional and normalize rendered evidence

🐞 Bug fix ⚙️ Configuration changes 🕐 20-40 Minutes

Grey Divider

AI Description

• Routes unstable L4 step attribution and evidence-mirror misses to optional findings.
• Preserves core failures for missing destinations and duplicate Garnet App comments.
• Normalizes rendered evidence and safely aggregates posture notes under nounset.
Diagram

graph TD
  C["App Comment"] --> D{"Duplicate?"} -->|"Yes"| F["Core Failure"]
  C --> N["Render Normalizer"] --> E{"Evidence present?"}
  E -->|"Destination missing"| F
  E -->|"Step missing"| O["Optional Report"]
  P["PR Description"] --> M{"Mirror current?"} -->|"No"| O
Loading
High-Level Assessment

The following are alternative approaches to this PR:

1. Configure tiers per L4 sub-check
  • ➕ Makes the tier configuration the executable source of truth.
  • ➕ Allows individual carve-outs to be promoted without editing shell routing.
  • ➕ Scales better if additional L4 exceptions are introduced.
  • ➖ Requires a tier-schema and reporting change for one leg's sub-checks.
  • ➖ Adds migration and naming complexity to existing leg identifiers.
  • ➖ Creates broader review scope for a limited rollout adjustment.

Recommendation: Keep the PR's scoped routing for this rollout because it preserves L4 as a core evidence leg while distinguishing its experimental checks. If more L4 carve-outs appear or these exceptions require promotion, introduce configurable sub-check tiers so policy is not duplicated between YAML prose and shell branches.

Files changed (2) +30 / -16

Bug fix (1) +29 / -15
garnet-release-gate-verify.shNormalize L4 evidence and route experimental misses as optional +29/-15

Normalize L4 evidence and route experimental misses as optional

• Normalizes defanged domains and numbered step names before checking rendered comment evidence. It routes missing quoted names and absent or stale description mirrors to optional findings while preserving core destination and duplicate-comment failures, and safely builds posture state from potentially empty arrays under nounset.

.github/scripts/garnet-release-gate-verify.sh

Other (1) +1 / -1
leg-tiers.ymlDocument L4's optional sub-check carve-outs +1/-1

Document L4's optional sub-check carve-outs

• Expands L4's tier rationale to identify quoted-step attribution and description-mirror misses as optional. It explicitly retains missing destinations and duplicate App comments as core failures.

.github/garnet-gate/leg-tiers.yml

@qodo-code-review

Copy link
Copy Markdown

Code review by qodo was updated up to the latest commit afc58c1

@qodo-code-review

Copy link
Copy Markdown

PR Code Suggestions ✨

Warning

/improve is deprecated. Use /agentic_review instead (removal date not yet scheduled).

No code suggestions found for the PR.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

garnet-release-testing Garnet product/release test PR; the Jibril release gate runs and reports here

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant