Repository navigation
cairn: install distilled knowledgebase from dd-backend-check investigations - #133
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1aa964bbc0
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| - **Before feature/design work on an area**: query it — run the cairn skill's | ||
| script: `python3 <cairn-skill-dir>/investigation.py brief <topic> [--paths <files>] [--diff]` (the cairn-memory skill — its directory is listed in your available skills; the `/cairn` plugin command wraps the same script on | ||
| Claude Code). Matched memories are design |
There was a problem hiding this comment.
When the external cairn-memory skill is not installed—as in a clean checkout and the current Codex environment—<cairn-skill-dir> is only a placeholder, no repository-local investigation.py exists, and no cairn executable is available. This makes the newly mandatory pre-design query workflow impossible to follow; include a repository-local helper or document a direct-search fallback for the committed memories.
AGENTS.md reference: AGENTS.md:L12-L15
Useful? React with 👍 / 👎.
| from 2fc44cc: `reports/l2-splitter-v3-dropped.patch` (v3 = final, includes both post-ship | ||
| fixes; v3 = v2 + 2 files, v2 = v1 + 3 files), `reports/l2-splitter-v1-benchmark.patch` |
There was a problem hiding this comment.
Following these revival instructions from the repository root fails because neither reports/l2-splitter-v3-dropped.patch nor reports/l2-splitter-v1-benchmark.patch exists there; the committed files are under doc/investigations/multi-64k/reports/. The same stale paths occur in the computeExtras and L2-verdict memories, so update them consistently before agents rely on this canonical revival procedure.
AGENTS.md reference: AGENTS.md:L6-L10
Useful? React with 👍 / 👎.
| - **L3 (last resort)**: `RuntimeCompiler.compile()` catches `MethodTooLargeException` → | ||
| `fallbackOrThrow` → JDK `java.util.regex` delegation with a warning (reggie-runtime, | ||
| ~line 1042). Comment names NFABytecodeGenerator as a generator without splitting. |
There was a problem hiding this comment.
For the default Reggie.compile() path, fallbackOrThrow does not delegate to java.util.regex: RuntimeCompiler.fallbackOrThrow throws UnsupportedPatternException unless ALLOW_JDK_FALLBACK is enabled. Describing L3 as unconditional JDK delegation gives future method-size work the wrong API and safety contract; state that it refuses in strict mode and delegates only with the opt-in option.
AGENTS.md reference: AGENTS.md:L6-L10
Useful? React with 👍 / 👎.
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #133 +/- ##
=========================================
- Coverage 85.2% 85.2% -0.1%
Complexity 1 1
=========================================
Files 163 163
Lines 48830 48830
Branches 6978 6978
=========================================
- Hits 41614 41612 -2
- Misses 5206 5207 +1
- Partials 2010 2011 +1 see 1 file with indirect coverage changes Continue to review full report in Codecov by Harness.
🚀 New features to boost your workflow:
|
2 participants
What does this PR do?
Installs a distilled cairn knowledgebase (
.knowledgebase/memories/, 37 memory units) from the frozen dd-backend-check investigations, plus the AGENTS.md instruction block that tells agents to consult it. No code changes.The memories cover: dd-backend regex adoption findings (prod CPU distribution, refusal-set parity, prefilter/scan arc results), JDK-compatibility fixes (NUL, bare
}, groupCount, CASE_INSENSITIVE, property classes), bounded-compile guarantees (work budget/deadline/emission cap), method-size layering (L1/L2/L3, incl. the dropped L2 splitter with revival patches), ASM/JVM codegen gotchas, and environment pitfalls (stale-jar trap, OpenJ9 deadline cliff, gradle caching).Motivation
Cross-investigation distillation (
cairn distill --all→ curated →cairn memorize), so future sessions load distilled, generally-valid knowledge instead of re-deriving it. Direct push to main is blocked by mergegate; this is the PR path.Related Issue(s)
None.
Change Type
Checklist
./gradlew build) — no code changed; tests unaffectedPerformance Impact
None — non-code content only.
Additional Notes
AI-assisted change (curated + installed via cairn-memory tooling).