docs(bench): reconcile §9.12 with the resolved #444 decision

[jensholdgaard]

What

Reconcile docs/benchmarks.md §9.12 with the now-resolved #444 decision. The per-service C2 gate shipped in #451 (option 3), but §9.12 still described #444 as an open three-way fork and an undecided "decision".

Change

• The kafka-fragmentation paragraph: "Tracked as miner: template fragmentation on short high-cardinality-token lines (kafka, v8 C2 finding) #444 (tokenizer masking vs. length-aware thresholding vs. accept-and-scope — an RFC-level pillar docs: apply RFC maturity-model amendments #2 decision)" → miner: template fragmentation on short high-cardinality-token lines (kafka, v8 C2 finding) #444 resolved (2026-07-10): option 3 shipped (per-service gate, RFC 0006 §3.4.3, PR feat(bench): redefine C2 gate as per-service (#444) #451); masking parked as a future strategic RFC (a Collector transform/redaction processor can polish high-cardinality infra tokens upstream); length-thresholding rejected.
• The B2-pricing closing line: "So miner: template fragmentation on short high-cardinality-token lines (kafka, v8 C2 finding) #444 is a query-capability / thesis-value decision" → the fragmentation is a query-capability tradeoff that miner: template fragmentation on short high-cardinality-token lines (kafka, v8 C2 finding) #444 accepted on hostile infra logs — the per-service gate makes C2 acceptance honest without masking.

Doc-only; no code, no gate numbers change (the v8 verdict already flipped FAIL→PASS in #451). mdbook build clean.

Hazard note

No pillar #2 behaviour changes — this is narrative reconciliation of a resolved decision, keeping the honest-measurement framing (per-service gate; masking left to the Collector per "format parsing is the Collector's job").

🤖 Generated with Claude Code

https://claude.ai/code/session_01X467Druw4cQeEPVp52DTG3

Summary by CodeRabbit

• Documentation
  • Updated benchmark narrative with the correct date-stamped resolution for per-service C2 gating.
  • Clarified that masking was deferred to a later upstream processing approach, and that length-aware thresholding was rejected.
  • Refined the hostile Kafka-shaped service discussion: pruning and latency remain unaffected, while template fragmentation reduces template-exact precision and limits what a single probe can recover.

[Copilot]

Pull request overview

Reconciles docs/benchmarks.md §9.12 with the now-resolved #444 decision, updating the narrative to reflect that option 3 (per-service C2 gate) has shipped (PR #451 / RFC 0006 §3.4.3) and reframing the kafka-driven fragmentation as an accepted query-capability tradeoff rather than an open “decision”.

Changes:

• Updates the §9.12 kafka-fragmentation discussion to explicitly mark #444 as resolved (2026-07-10) and document the chosen path (per-service C2 gate), with masking deferred and length-aware thresholding rejected.
• Adjusts the B2 pricing conclusion to describe fragmentation as an accepted query-capability / thesis-value tradeoff on hostile infra logs, not a performance issue.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[coderabbitai]

Warning

Review limit reached

@jensholdgaard, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 24 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 0193e533-5855-4793-9700-dbf0a2e73ce8

📥 Commits

Reviewing files that changed from the base of the PR and between 03b9aa1 and 54fa5ea.

📒 Files selected for processing (1)

• docs/benchmarks.md

📝 Walkthrough

Walkthrough

Section 9.12 documents RFC #444 as resolved, records the selected and deferred approaches, and clarifies that template fragmentation affects template-exact precision without changing pruning or latency floors.

Changes

C2 benchmark documentation

Layer / File(s)	Summary
Resolved RFC decision and fragmentation tradeoff docs/benchmarks.md	Section 9.12 adds dated C2 rescoring, records per-service gating, deferred masking, rejected length-thresholding, preserved safety constraints, and fragmentation’s effect on template-exact recovery while pruning and latency remain unchanged.

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

Possibly related PRs

• jensholdgaard/ourios#445: Defines the related per-service C2 decomposition and fragmentation tradeoffs.
• jensholdgaard/ourios#451: Redefines the per-service C2 gate semantics documented here.
• jensholdgaard/ourios#48: Specifies C2 metric semantics and benchmark documentation updates in Section 9.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title is concise and accurately reflects the doc-only reconciliation of §9.12 with the resolved #444 decision.
Description check	✅ Passed	It explains the purpose, changes, and related issue context, though it doesn't follow the template headings exactly.
Linked Issues check	✅ Passed	It implements the #444 resolution by documenting option 3, masking deferral, length-threshold rejection, and the accepted tradeoff.
Out of Scope Changes check	✅ Passed	The change stays within docs/benchmarks.md and only reconciles §9.12 wording; no unrelated code or scope was added.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch docs-c2-9-12-reconcile

---

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

• X
• Mastodon
• Reddit
• LinkedIn

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

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@docs/benchmarks.md`:
- Around line 1148-1155: Align the Section 9.12 heading date with the July 10,
2026 decision described in the text, or explicitly clarify in the section that
the decision and RFC 0006 amendment occurred after the July 9 benchmark run.
- Around line 1186-1192: Clarify the benchmark statement near the “~11,651 ids”
discussion: replace “no single template query recovers it” with wording that
acknowledges partial recovery while stating that no single template query
recovers the full dominant event family, consistent with the measured 11,523
recovered rows.

🪄 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: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 8b939e62-268e-4165-aba4-0c2fa79098d6

📥 Commits

Reviewing files that changed from the base of the PR and between b257c9a and fdd6999.

📒 Files selected for processing (1)

• docs/benchmarks.md

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.