Repository navigation
Security, smell and test-speed cards; split large indexes - #30
Merged
Merged
Conversation
A level index past its budget now gets a short hub page that links its context parts, so llms.txt keeps one index link per level. The budget counts the frontmatter the site adds. orgs.py gains the organisations and URL prefixes the new cards cite. AUDIT_STEPS gains four lines (hardening is not a FAIL, say when partial, adversarial and empirical modes). The header nav is a CSS-only dropdown. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013wy5PHVgonRFe4D63zkzxr
New cards from the Cloudflare security-audit classes: request framing, cache keys, Host trust, header CRLF, OAuth flow binding, SAML, MFA step-up, account linking, password reset, API key scope, soft delete, export scope, tenant id in telemetry, decompression cap, webhook replay window and workload identity. Clauses added to session-lifecycle, api-cache-control, dlq, rate-limit-quota and workflow-supply-chain. Two models judged each card; fixes applied from docs/research/judge-triage-2026-11-06.md. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013wy5PHVgonRFe4D63zkzxr
Fixtures built once, suite time budget, test selection on pull requests, pytest fixture scope, xdist worker isolation, collection and import time, and test hygiene. Adds a draft for committed per-user IDE and OS files. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013wy5PHVgonRFe4D63zkzxr
Every remaining Fowler and Pocock smell is now a card or merged into one: length and parameter limits, class size, caller mutation, import cycles, fan-out, Law of Demeter, shotgun surgery, repeated switches, temporary field, and merges into existing cards. Several drafts promoted where three organisations back each Rule clause. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013wy5PHVgonRFe4D63zkzxr
Output of tools/build.py and lint/scripts/snapshot.py. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013wy5PHVgonRFe4D63zkzxr
Source searches, spot checks, evidence files, judge reports and the triage behind this batch of cards. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013wy5PHVgonRFe4D63zkzxr
A level page whose text passes PAGE_MAX (now in tools/config.py) becomes a compact hub that keeps a row and id for every card, plus one full page per context with its own Markdown twin. One helper decides which page holds a card, and the context grouping is shared with the index split. service.html goes from 134,780 to 15,090 characters. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013wy5PHVgonRFe4D63zkzxr
Drop a duplicate "ecto" org key, use real annotations with postponed evaluation, wrap a long line, and give mypy the Level type it expects. The golden hashes for /concerns.md are recaptured because this batch changes the stable card set. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013wy5PHVgonRFe4D63zkzxr
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
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 batch adds 42 new or edited cards (35 stable, 7 draft) plus the source research behind them.
context, sollms.txtkeeps one index link per level. The function and service indexes split this way now.Every stable card needs three organisations behind each Rule clause. Two models (Opus and Sonnet) judged each card on its own, and the fixes follow
docs/research/judge-triage-2026-11-06.md. Spot checks of the research quotes are indocs/research/spot-checks-2026-11-06.md.This is one PR for many cards, against the one-card-per-PR rule, by the maintainer's choice.
PAGE_MAX, now intools/config.py) becomes a compact hub that keeps a row and id for every card, plus one full page per context with its own Markdown twin.service.htmlgoes from 134,780 to 15,090 characters, so oldservice.html#<id>links still land.pytest shows the same 5 failures as
main(test_sbomand the ruff-dependent lint tests).🤖 Generated with Claude Code
https://claude.ai/code/session_013wy5PHVgonRFe4D63zkzxr