Implementation guide: derive unfinished-loops query from recent conversation topics - #107
Draft
Dithilli wants to merge 3 commits into
Draft
Implementation guide: derive unfinished-loops query from recent conversation topics#107Dithilli wants to merge 3 commits into
Dithilli wants to merge 3 commits into
Conversation
…roposal/04 sequencing disagreement for human decision
Dithilli
added a commit
that referenced
this pull request
Jul 13, 2026
… 4, PR #87) Phase 0 of proposal 04, reconciled to post-0.19.0 main. Read-only audit of the startup-orientation loops block: - scripts/audit-loops.ts: owner-run live runner that calls the REAL gatherOrientation (extracted in #111) for a byte-identical as-injected block, plus the hook's exact underlying client.search call for per-result diagnostics; deterministic JSONL for a diffable daily ledger; --simulate a1,a2,dynamic,after30,after60 compares proposal-04 candidates and PR #107's dynamic-query shape side by side without touching production behavior. - lib/loops-audit.ts + lib/loops-audit.test.ts: pure parsing/formatting/ summary helpers, hermetically tested in npm test (the live layer never is), mirroring The Ruler's scripts/eval-retrieval.ts + lib/eval-matchers.ts split. - docs/loops-audit-rubric.md: judging rubric, metrics, decision rule with remedy tiers (tune static string → #107 dynamic query → explicit tagging), and the two-week owner workflow. - Content in output is opt-in via HYPERSPELL_AUDIT_CONTENT=1 with delete-after-window, per the HYPERSPELL_SCORE_LOG precedent; the ledger directory is gitignored. Zero production change: gatherOrientation, the handler, and the default loopsQuery are untouched; all existing tests pass unmodified. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016M7SehQ6Ysh7Cvs2oEYgcG
Dithilli
added a commit
that referenced
this pull request
Jul 13, 2026
… 4, PR #87) Phase 0 of proposal 04, reconciled to post-0.19.0 main. Read-only audit of the startup-orientation loops block: - scripts/audit-loops.ts: owner-run live runner that calls the REAL gatherOrientation (extracted in #111) for a byte-identical as-injected block, plus the hook's exact underlying client.search call for per-result diagnostics; deterministic JSONL for a diffable daily ledger; --simulate a1,a2,dynamic,after30,after60 compares proposal-04 candidates and PR #107's dynamic-query shape side by side without touching production behavior. - lib/loops-audit.ts + lib/loops-audit.test.ts: pure parsing/formatting/ summary helpers, hermetically tested in npm test (the live layer never is), mirroring The Ruler's scripts/eval-retrieval.ts + lib/eval-matchers.ts split. - docs/loops-audit-rubric.md: judging rubric, metrics, decision rule with remedy tiers (tune static string → #107 dynamic query → explicit tagging), and the two-week owner workflow. - Content in output is opt-in via HYPERSPELL_AUDIT_CONTENT=1 with delete-after-window, per the HYPERSPELL_SCORE_LOG precedent; the ledger directory is gitignored. Zero production change: gatherOrientation, the handler, and the default loopsQuery are untouched; all existing tests pass unmodified. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016M7SehQ6Ysh7Cvs2oEYgcG
Dithilli
added a commit
that referenced
this pull request
Jul 13, 2026
…udit-first (#87) * docs(proposal): startup-orientation loops tuning (idea #4 from #66) Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> * feat(audit): unfinished-loops audit tooling — Loose Threads (#66 idea 4, PR #87) Phase 0 of proposal 04, reconciled to post-0.19.0 main. Read-only audit of the startup-orientation loops block: - scripts/audit-loops.ts: owner-run live runner that calls the REAL gatherOrientation (extracted in #111) for a byte-identical as-injected block, plus the hook's exact underlying client.search call for per-result diagnostics; deterministic JSONL for a diffable daily ledger; --simulate a1,a2,dynamic,after30,after60 compares proposal-04 candidates and PR #107's dynamic-query shape side by side without touching production behavior. - lib/loops-audit.ts + lib/loops-audit.test.ts: pure parsing/formatting/ summary helpers, hermetically tested in npm test (the live layer never is), mirroring The Ruler's scripts/eval-retrieval.ts + lib/eval-matchers.ts split. - docs/loops-audit-rubric.md: judging rubric, metrics, decision rule with remedy tiers (tune static string → #107 dynamic query → explicit tagging), and the two-week owner workflow. - Content in output is opt-in via HYPERSPELL_AUDIT_CONTENT=1 with delete-after-window, per the HYPERSPELL_SCORE_LOG precedent; the ledger directory is gitignored. Zero production change: gatherOrientation, the handler, and the default loopsQuery are untouched; all existing tests pass unmodified. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016M7SehQ6Ysh7Cvs2oEYgcG --------- Co-authored-by: Claude Sonnet 5 <noreply@anthropic.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.
Detailed implementation guide for #73. See
docs/plans/issue-73-static-loops-query.md— appends recent conversation titles to the static intent phrase so the loops query adapts per-session, with a documented (out of scope) alternative for explicit loop tagging.Ref #73 — planning/design doc, not the implementation.