Skip to content

canon: canon-integration-audit — three checks the existing gauntlets don't catch (follow-up to #134)#135

Merged
klappy merged 2 commits intomainfrom
feat/canon-integration-audit
Apr 26, 2026
Merged

canon: canon-integration-audit — three checks the existing gauntlets don't catch (follow-up to #134)#135
klappy merged 2 commits intomainfrom
feat/canon-integration-audit

Conversation

@klappy
Copy link
Copy Markdown
Owner

@klappy klappy commented Apr 24, 2026

What and why

Follow-up to #134. Codifies the three integration-layer gaps surfaced during the dream-house essay drafting session (rev 0-23):

Audit What it checks Fires at Closes gap from PR #134
Concept audit Every named concept intended for reader adoption has a canon home or explicit essay-only decision on Dream-house-principle / discernment-layer were essay-only until rev 23
Adjacent-canon audit Every load-bearing claim cites all matching existing canon via capability-is-not-permission + mode-discipline-bottleneck-respect weren't cited until rev 23
Validator-completeness audit Same-session validators implement full schema type-discipline, not just presence Rev-21 list-vs-string violation false-green caught only by MA re-validation

Why one doc, not three

All three gaps share a root cause: existing gauntlets optimize for is this document well-written? not does this document integrate with the rest of the knowledge base? One tier-2 constraint with three named sub-audits:

  • Shared root cause → shared fix
  • Lower drift surface (one source of truth)
  • Easier extension when a fourth integration check emerges
  • One URI citable from preflight/gate/validate

Honest scope note — canon is content, tool-wiring is trigger

This doc defines what the audits check. The proactive-enforcement side depends on oddkit tools exposing the audits at the named transition points. That's in the klappy/oddkit repo, not this one — extends P11 (which currently tracks mechanical enforcement of alone).

Until tool wiring ships, this canon is passively discoverable via and citable from existing workflow, but the audits do not fire without invocation. Both pieces (canon + tool) are required for the proactive property.

Files changed

  • NEW (tier 2, 3,156 words, 10 sections)
  • — See Also updated (reciprocal ref)
  • — See Also updated (reciprocal ref)

Pre-merge validation

  • Local frontmatter type-discipline check passes (tier: int, derives_from: str, all universal 8 present)
  • Managed Agent independent validator (dispatching after PR creation)
  • Cursor Bugbot

Note

Low Risk
Low risk: documentation-only changes that add a new governance constraint and update a few cross-references; no runtime code or schema behavior changes.

Overview
Introduces canon/constraints/canon-integration-audit.md, a new tier-2 constraint that formalizes three pre-merge integration checks: concept canonization, adjacent-canon citation completeness, and frontmatter validator completeness (type-discipline), with intended trigger points at oddkit_preflight, oddkit_gate, and oddkit_validate.

Updates See Also cross-references in frontmatter-validation-before-merge.md and revision-lens-sequence.md, and adds reciprocal links from dream-house-principle.md and discernment-layer.md back to the new audit constraint.

Reviewed by Cursor Bugbot for commit 66d4203. Bugbot is set up for automated code reviews on this repo. Configure here.

klappy added 2 commits April 24, 2026 03:23
…don't catch

Follow-up to PR #134. Codifies the three integration-layer gaps
surfaced during the dream-house essay drafting session (rev 0-23):

  1. Concept audit — named concepts intended for reader adoption
     must have a canon home, or an explicit essay-only decision
     must be recorded. (Rev-5 'Dream House Principle' concept
     would have been canonized here instead of rev 23.)

  2. Adjacent-canon audit — load-bearing claims must cite all
     matching existing canon via derives_from. (Rev-0 derives_from
     would have included capability-is-not-permission and
     mode-discipline-and-bottleneck-respect instead of rev 23.)

  3. Validator-completeness audit — same-session validators whose
     green verdict closes the merge gate must implement full schema
     type-discipline, not just presence. (Rev-21 list-vs-string
     violation would have been caught same-session instead of
     round-tripping through the MA validator at rev 22.)

Structured as ONE tier-2 constraint rather than three smaller docs:

  - Shared root cause (extrinsic integration missing from gauntlet)
  - Lower drift surface (one source of truth)
  - Easier extension when a fourth integration check emerges
  - Cross-references stay tight (one URI referenced from preflight,
    gate, validate instead of three)

Trigger points specified for each audit:

  adjacent-canon audit    → oddkit_preflight
  concept audit           → oddkit_gate (drafting → peer-review-ready)
  validator-completeness  → oddkit_validate

HONEST SCOPE NOTE — canon is content, tool-wiring is trigger:

  This doc defines what the audits check. The proactive-enforcement
  side depends on oddkit tools exposing the audits at the named
  transition points. That work is in the klappy/oddkit repo, not
  this one — it extends P11 (which currently tracks mechanical
  enforcement of release-validation-gate alone).

  Until the tool wiring ships, this canon is passively discoverable
  via oddkit_search and citable from the existing workflow, but the
  audits do not fire without the author or agent invoking them.
  Both pieces (canon + tool) are required for proactive property.

Reciprocal See Also entries added to:

  - canon/methods/revision-lens-sequence (the essay workflow this
    audit extends)
  - canon/constraints/frontmatter-validation-before-merge (the
    narrower rule the validator-completeness audit extends)

Cross-references verified complete in both directions: the audit
references revision-lens-sequence, DoD, release-validation-gate,
frontmatter-schema, DRY-canon-says-it-once, dream-house-principle,
discernment-layer, axioms, and the dream-house essay. Each referenced
doc either already references the audit back (via these See Also
updates) or is a canonical foundation that does not need back-refs.
MA validator PASSED PR #135 with one non-blocking observation: the
two complemented docs (dream-house-principle, discernment-layer)
lacked reciprocal back-links to canon-integration-audit.

The audit itself prescribes that cross-reference completeness be
bi-directional for docs that are genuinely related. Since both
principles are explicitly named in the audit's Application section
as 'the canon docs whose absence during the dream-house essay
drafting motivated the concept audit', the back-reference belongs.

Closing this within the same PR rather than as follow-up — the fix
is two one-line See Also additions and avoids shipping a known
asymmetry that the audit itself flags.
@klappy klappy merged commit 365b553 into main Apr 26, 2026
1 check passed
@klappy klappy deleted the feat/canon-integration-audit branch April 26, 2026 02:53
klappy added a commit that referenced this pull request Apr 26, 2026
… DOLCHEO artifacts) (#138)

Captures the four-PR session per the milestone journaling gate:

  PR #134 — Penny Wise and Pound Foolish essay + 4 canon docs
  PR #135 — canon-integration-audit constraint
  PR #137 — telemetry semantic names interface (oddkit)
  PR #138 — cache_tier streaming-race fix (oddkit)

14 DOLCHEO artifacts: 3 D, 2 O, 3 L, 2 C, 2 O-open, 1 H.

Encode does not persist; this is the file form.
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.

1 participant