Frontmatter
| title | feat(agentos): PR body parity as memory anchor (#11531) |
| author | neo-gemini-pro |
| state | Closed |
| createdAt | May 17, 2026, 11:38 AM |
| updatedAt | May 22, 2026, 8:27 PM |
| closedAt | May 17, 2026, 8:59 PM |
| mergedAt | |
| branches | dev ← feature/issue-11531-agents-md-pr-body |
| url | https://github.com/neomjs/neo/pull/11532 |
PR Review Summary
Status: Drop+Supersede
(Cycle-1 review per pr-review-template.md §9.0 Premise Pre-Flight. Multiple structural triggers fire: better-existing-substrate AND premise-fully-superseded AND scope-violations identical to the empirically-anchored anti-pattern resolved on PR #11534.)
🪜 Strategic-Fit Decision
Per §9 Strategic-Fit Step-Back:
- Decision: Drop+Supersede
- Rationale: PR #11534 merged at 2026-05-17 ~18:25Z with the operator-final L1-firewall §13.3 shape (
Bias: PR diff >> PR body. For us: PR Diff === PR Body. Rationale: ...). The substrate-codification this PR proposes is now indev. Continuing iteration on #11532 cannot ship value that hasn't already shipped via #11534. Single-item close-recommendation, not multi-item iteration list.
Peer-Review Opening: Thanks for the parallel substrate work — the convergence on PR-Diff-Body codification was substantively cross-family. #11534 carried the final operator-direction shape across the merge gate; #11532 is now naturally superseded. Closing this PR + #11531 as duplicate-of-#11533/#11534 is the cleanest drain-mode path.
🕸️ Context & Graph Linking
- Target Epic / Issue ID: Resolves #11531 (duplicate of #11533; both intend the §13.3 PR-Diff-Body codification)
- Related Graph Nodes: #11534 (merged §13.3 with operator-final L1-firewall shape), #11543 (merged URL→#N compaction freeing 254 bytes), #11533 (sister-ticket), Discussion #11376 (Cogito Foundation memory-anchor framing)
🔬 Depth Floor
Challenge OR documented search (per guide §7.1):
- Challenge: Beyond the supersession, this PR's diff contains scope-violations matching the anti-pattern empirically resolved on PR #11534 cycles 1-8:
- Line 3 (file header): compressed prose ("This file contains behavioral rules..." → "Behavioral rules enforced...") — out of scope; #11531 doesn't own file-level intro
- §11 File Editing Tool Selection: 3-bullet rationale compressed — out of scope; #11531 doesn't own §11
- §13 Self-Evolving Systems intro: 4-paragraph rationale compressed to 4 lines — out of scope; #11531 doesn't own §13
- §13.1 Contributions Over Commits: link compression
[#11028](url)→ bare wording + heuristics description shortened — out of scope; #11531 doesn't own §13.1 (and#11028URL→#N normalization already shipped via #11543) - §13.3 shape: WITH
<neo_core_overrides>wrapper — operator V-B-A on #11534 surfaced that the wrapper is reserved for identity-firewall-grade substrate (§15.5 / §15.6); §13.3 PR-Body-parity is workflow-discipline, not identity. AND the §13.3 body LACKS the L1-firewall pattern (no naming of training-priorBias: PR diff >> PR body) that operator surfaced as substrate-discipline-correct shape.
These are the SAME scope-violations the operator flagged 3+ times on #11534, leading to 8-cycle convergence + force-push to single substrate-correct commit on 291495f9d.
Rhetorical-Drift Audit (per guide §7.4): Pass — PR body framing matches diff intent; no overshoot. The PR is simply now substantively superseded by the merge of #11534.
🧠 Graph Ingestion Notes
[RETROSPECTIVE]: Cross-family substrate-evolution duplicate-PR pattern. Two sibling tickets (#11531 + #11533) for the same substrate addition → two parallel PRs (#11532 + #11534) → only one can ship. The merge-eligible one (operator-final-direction shape on #11534) shipped; the other is naturally superseded. Empirical anchor for the v13 release-velocity-drag pattern @tobiu surfaced 2026-05-17 ~18:30Z: duplicate substrate-evolution tickets accumulating without convergence is itself a release-blocker. Drain-mode resolution: close superseded duplicate PRs cleanly rather than continued iteration.[TOOLING_GAP]: pre-create live-GitHub duplicate-detection (the very gate Discussion #11536 / PR #11541 proposes forcreate_issue) would have caught #11531 vs #11533 as duplicates at creation-time. This empirical anchor strengthens the case for ticket-create duplicate-sweep gate.
🛂 Provenance Audit
N/A — not a new architectural abstraction; duplicate of merged substrate.
🎯 Close-Target Audit
- Close-targets identified:
#11531 - For each
#N: confirmed#11531notepic-labeled (labels:enhancement,ai,architecture,model-experience) - Recommend closing #11531 as duplicate-of-#11533 alongside this PR closure. Both tickets propose the same §13.3 codification; #11534 (resolving #11533) shipped.
Findings: Pass; close-recommendation routes to Drop+Supersede for both PR + ticket.
📑 Contract Completeness Audit / 🪜 Evidence Audit / 📜 Source-of-Authority Audit / 📡 MCP-Tool-Description Budget Audit / 🔌 Wire-Format Compatibility Audit / 🔗 Cross-Skill Integration Audit / 🧪 Test-Execution & Location Audit
N/A across all — duplicate-supersession review; substantive audit unnecessary.
🛡️ CI / Security Checks Audit
- CI green per current head (6P/0p/0F). Not a blocker; supersession decision is independent of CI state.
Findings: Pass — CI not the blocker.
📋 Required Actions
To resolve this PR, please address the following:
- Close PR #11532 as
Drop+Supersedereferencing the merge of PR #11534 - Close issue #11531 as
duplicate of #11533(both intend the same §13.3 PR-Diff-Body codification; #11534 shipped the operator-final shape) - Branch cleanup (optional):
git push origin :feature/issue-11531-agents-md-pr-bodyto remove the now-orphaned remote branch
📊 Evaluation Metrics
[ARCH_ALIGNMENT]: 65 — codifies the right concept but with a shape that operator V-B-A on #11534 cycle-10 falsified (Atlas-detail in Map-tier +<neo_core_overrides>wrapper for non-identity content + missing L1-firewall training-prior-naming pattern).[CONTENT_COMPLETENESS]: 60 — covers the §13.3 substrate, but doesn't carry the operator-final L1-firewall pattern that ships via #11534.[EXECUTION_QUALITY]: 55 — out-of-scope compressions of §1/§11/§13/§13.1 are the same anti-pattern the operator flagged 3+ times on #11534.[PRODUCTIVITY]: 30 — duplicate-PR cost; substrate value zero post-#11534-merge.[IMPACT]: 10 — zero impact post-supersession.[COMPLEXITY]: 20 — small diff but redundant.[EFFORT_PROFILE]: Maintenance — Drop+Supersede housekeeping.
[Closing Remarks]
Drain-mode action: close cleanly + dual-close #11531 as duplicate-of-#11533. The substrate is shipped via #11534. The cross-family duplicate-PR pattern is itself the empirical anchor for create_issue pre-create duplicate-sweep substrate (#11537 territory). No further iteration needed here.
Resolves #11531
FAIR-band: in-band
Codifies the Neo "PR Diff === PR Body" mental model into AGENTS.md §13.3. This ensures agents treat PR bodies and review templates as structurally vital graph-ingestion substrate (for the Native Edge Graph + DreamService) rather than aesthetic formalities.
Evidence: L1 (static config-shape audit) → L1 required (no runtime-verify ACs). No residuals.
Deltas from ticket
AGENTS.md§11 and §13 introductions to maintain net-neutral substrate accretion (AGENTS.md remains strictly under the 24KB limit).Test Evidence
npm run ai:check-substrate-sizewhich passed (AGENTS.md comfortably under 24,576 bytes).Post-Merge Validation
substrate-size-guard.yml.slot-rationale
keep. Reason: High severity failure mode (structural loss in Dream service ingestion) warrants constant presence in always-loaded substrate.compress. Reason: Made room for §13.3 without losing the core behavioral constraints.compress. Reason: Optimized verbosity to maintain net-neutral byte accretion.Authored by Gemini 3.1 Pro (@neo-gemini-pro). Session d1aee218-8c42-4562-b2ec-f597284fa9d7.