[DRAFT] docs: PR #387 Kiro-vs-Claude evaluation + reconciliation plan (independent Linux/Py3.12 run)#16
Open
plauzy wants to merge 1 commit into
Conversation
…ion plan Evidence-based head-to-head of the Kiro (#13/#14) and Claude (#11/#12) remediations of upstream awslabs#387: executed gate results on all four branches (Python 3.12.13), a per-finding scorecard, a 'what both missed' list, and a step-by-step reconciliation plan for two reconcile/* branches. Plan only — no reconciliation code. Co-authored-by: Kiro Agent <244629292+kiro-agent@users.noreply.github.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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This pull request was created by @kiro-agent on behalf of @plauzy 👻
Comment with /kiro fix to address specific feedback or /kiro all to address everything.
Learn about Kiro Web
What this is
An independent, evidence-based head-to-head evaluation of the two remediations of upstream awslabs/cli-agent-orchestrator#387, plus a reconciliation plan. Adds one file:
docs/reviews/pr387-reconciliation-plan.md. Plan only — no reconciliation code.kiro/pr387-agui-core, PR-B #13kiro/pr387-a2a-hardenedclaude/pr387-agui-core, PR-B #12claude/pr387-a2a-hardenedWhy a second branch
A concurrent run already pushed
docs/pr387-reconciliation(a complete plan). This PR is a separate, independently-derived pass on a different toolchain, deliberately not clobbering that branch. The two can be diffed against each other. Base isfeat/agentic-protocols-generative-ui; please mark as draft (the create step could not set the draft flag directly).How it was verified (this run)
All four branches + base fetched as source; every gate executed on Linux, Python 3.12.13,
uv sync --all-extras --dev, fullpytest(including e2e):kiro/pr387-agui-coreclaude/pr387-agui-core[otel]subprocess test uses the legacyfind_modulehook, ignored on Py≥3.12)kiro/pr387-a2a-hardenedclaude/pr387-a2a-hardened(Note: this differs from the macOS/
-m 'not e2e'concurrent run, which attributed a failure tokiro-a2aand none toclaude-a2a— the flakiness is environment/selection-sensitive, which the plan documents.)Highlights of the verdict
?since=cursor-loss fix (Kiro'sapi.tsis byte-identical to base), higher-fidelity live reconnect (real server SIGKILL→replay), real-JWT/JWKS auth tests + thereset_jwks_cachelatent-break fix, and enforcement-before-parse.SHIPPED_SKILLS), endpoint-replay depth, the extracted testable_should_mount_a2a()guard, and the correct "no false[a2a]extra" dependency call.emit_uirate limiting,STATE_DELTAdebounce.Full scorecard, "what both missed", and the per-contrast reconciliation decisions (two
reconcile/*branches, gates, constraints, updated upstream reply, retarget sequencing) are in the file.