Skip to content

CoalBoard v1.3.0 - layman-default wizard + flow/honesty rework

Choose a tag to compare

View all tags
@HetCreep HetCreep released this 20 Jun 22:01
· 15 commits to main since this release
v1.3.0
This tag was signed with the committer’s verified signature.
HetCreep
SSH Key Fingerprint: 0osv9sipbbl2qmrI00rRA3kcEKs4ZuS+FBihPD/c8DU
Verified
Learn about vigilant mode.
42dfc58
This commit was signed with the committer’s verified signature.
HetCreep
SSH Key Fingerprint: 0osv9sipbbl2qmrI00rRA3kcEKs4ZuS+FBihPD/c8DU
Verified
Learn about vigilant mode.

Round-3 deep dogfood — the user ran the real published wizard + board as a customer and surfaced flow / honesty gaps the build gate cannot. MINOR: a new layman-default UX + a holistic wizard/gate flow rework + model-diversity honesty; no new config keys.

Added

  • Dual-audience wizard — a LAYMAN-DEFAULT path: /coalboard now defaults to AI-handles-everything — smart safe defaults (cwd · auto-work-type · L2 · standard) + ONE plain-language bill+confirm ("3 reviewers + a judge check X for ~Y tokens — go / cheaper / more thorough / cancel"; no opaque jargon — "cheaper/more thorough" map to depth/rigor, a universally-known word like "nasa" may stay). A programmer opts into the full restaurant wizard. The layman is kept safe by staging + the human-apply gate; the result carries the honest ceiling in plain language and never says "definitely safe".

Changed

  • Wizard = the "restaurant" order → bill → pay (programmer path): TARGET → silent scan → the 3 settings → the ACCURATE bill computed FROM the picks → ONE confirm. Fixes the stale-cost consent (the bill used to precede the picks).
  • Step 4 exit re-ordered + leaned: DIGEST → ONE consent question (apply-all / let-me-pick / report-only / stop) → THEN write, CONDITIONAL on the choice (stop = write NOTHING; the report is never written before consent).
  • Surfaced output = decisions + results only — internal mechanics (template fill, memory arm, contract steps) run SILENTLY.
  • Model-diversity honesty: diversifyModels is INERT on Claude Code (the spawn tool takes only aliases — it cannot pin a model generation). The only actuatable model-decorrelation is a tier-mix (partial, at a lens-strength cost); the real decorrelation is the diverse lens prompts + adversary + sub4, never the model.
  • NASA honesty (correlated-blind-spot ceiling): all-opus at nasa = MAX model-correlation at MAX stakes — the escape is the non-model ground-truth gates (tier2Verify) + the human, NOT model-diversity and NOT sub4 (sub4 is the same model → shares the blind spot; it breaks deadlocks only).
  • Warm-resume corrected (verified): the standard CC session has no callable SendMessage tool → recover a stopped/dead lens by re-spawning a FRESH lens on the un-done REMAINDER from main's journal; SendMessage-resume is a bonus only where the tool exists. TaskStop reaps a runaway/zombie sub.

Fixed

  • The org .github repo had no dependabot.yml — its workflows' pinned actions never auto-bumped; added it (github-actions, weekly), matching the plugin repos.

Gate: build + verify + 28 tests PASS.

Assets 2
Loading