Skip to content

2.0.1 — exhaust-audit fixes and fresh README graphics - #9

Merged
creativeskyai merged 2 commits into
mainfrom
claude/fable-opus5-overhaul-1i7tmj
Aug 8, 2026
Merged

creativeskyai merged 2 commits into
mainfrom
claude/fable-opus5-overhaul-1i7tmj

Conversation

@creativeskyai

Copy link
Copy Markdown
Owner

What changed

Fixes from running the pack's own fable-exhaust over v2.0.0 immediately after merge, plus the fresh README graphics.

From the audit (each finding survived a three-lens skeptic panel):

  • Installers — deleting .claude/settings.json is now a real opt-out: neither --update nor a plain re-run on an existing install re-creates it (materialization is keyed on the pre-existing VERSION stamp, so it is truly first-install-only). CI asserts the opt-out on Linux and Windows.
  • fable-exhaust — crash guards for a finding missing title/file and for a lost skeptic panel (drops are announced, never silent); rounds where every finder failed no longer count toward "ran dry" (a round nobody searched is not evidence of coverage — this exact failure occurred in the audit run itself); the 200-location dedup-hint truncation is announced; the skeptic panel is votes-sized, so fleet max now actually runs 5 skeptics per finding as CONFIG.md claims.
  • fable-review — vote count clamped to ≥ 1; a zero/negative votes would have spawned no skeptics and marked every raw finding as verified.
  • fable-migrate — the pilot gate fails closed when a pilot transform/check never completes.
  • fable-design — an all-judges-failed panel errors instead of silently crowning approach 0.
  • Docs — fable-ship's description matches its real gate ordering (read-only gates in parallel, build/tests after, alone); the /fable-config recreation template carries the full shipped key set.

Graphics: new SVG set — value-dial.svg (the two-dials config model), review-pipeline.svg (replaces the mermaid block), demo-run.svg (the real 2.0 dogfood review: 4 finders, 5 raw findings, 15 skeptic votes, 5 confirmed and fixed) — plus a regenerated 1280×640 social preview; the stale 1.0 screenshot is removed.

Why

The 2.0.0 release merged with its own audit tooling available, so we pointed it at itself. One finder round completed fully (4 finders, 13 raw findings verified by skeptic panels) before later rounds hit a usage limit — a bounded hunt, disclosed as such in the changelog. The confirmed findings were all real: two crash paths that lose accumulated results, two fail-open verification gates, one broken documented opt-out, and doc/code mismatches.

How it was verified

  • node tools/check-workflows.cjs passes (config-block and run() twins intact)
  • Installer scenarios pass locally: fresh install creates settings.json; plain re-run and --update both respect a deleted one; --update preserves user-edited CONFIG.md and settings.json
  • New CI step asserts the opt-out on both platforms
  • CHANGELOG.md 2.0.1 entry matches VERSION (checker-enforced)

🤖 Generated with Claude Code

https://claude.ai/code/session_01XWSx33Mi2bXpR6w75r3SY2


Generated by Claude Code

claude added 2 commits August 8, 2026 02:46
New SVGs: value-dial (the two-dials config model), review-pipeline
(replaces the mermaid block), and demo-run (the real 2.0 dogfood
review: 4 finders, 5 raw findings, 15 skeptic votes, 5 confirmed and
fixed). Social preview regenerated at 1280x640 with 2.0 framing; the
stale 1.0 demo screenshot is removed.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XWSx33Mi2bXpR6w75r3SY2
Confirmed by three-lens skeptic panels over one fully verified finder
round (later rounds were cut off by a usage limit — bounded hunt, said
so in the changelog):

- installers: deleting .claude/settings.json is a real opt-out now —
  no re-creation on --update or plain re-runs (first-install-only,
  keyed on the VERSION stamp; CI asserts it on both platforms)
- exhaust: crash guards for malformed findings and lost panels, failed
  finder rounds no longer count as dry, dedup-hint truncation announced,
  panel is votes-sized so fleet max really means 5 skeptics
- review: votes clamped >= 1 (zero skeptics used to verify everything)
- migrate: pilot gate fails closed on incomplete pilot results
- design: all-judges-failed now errors instead of crowning approach 0
- docs: ship gate ordering corrected; config template completed

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XWSx33Mi2bXpR6w75r3SY2
@creativeskyai
creativeskyai marked this pull request as ready for review August 8, 2026 13:19
@cursor

cursor Bot commented Aug 8, 2026

Copy link
Copy Markdown

Bugbot is not enabled for your account, so this pull request was not reviewed.

Enable Bugbot in the Cursor dashboard to get automatic reviews on future PRs.

@creativeskyai
creativeskyai merged commit 2f346a3 into main Aug 8, 2026
3 checks passed
@creativeskyai
creativeskyai deleted the claude/fable-opus5-overhaul-1i7tmj branch August 8, 2026 21:53
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants