Cross-Pollination Brief — April 21, 2026
Note: PM filed a retrospective omnibus log for April 16 on April 19. Most of its content — ethics activation, M2c closure, DECISIONS.md, the CXO voice guidance — is already reported in the April 16–18 briefs. This brief extracts the one item from that omnibus that was not visible in the atomic session logs at the time: a new canonical-term discipline step in Docs's omnibus skill.
Key Insights
1. PDR-004 correction chain — canonical-term discipline as process infrastructure
From: PM docs/omnibus-logs/2026-04-16-omnibus-log.md (filed April 19)
Relevant to: Klatch (agent process discipline, omnibus methodology)
CXO spotted that a paraphrase in an earlier PM omnibus had diverged from the canonical PDR-004 principles. Docs traced the drift: 7 affected files, 2 still live in published blog posts (Closing Sprint and Ship #036). Comms had to write full narrative rewrites — not find-and-replace — because the wrong principles had been tied to specific design decisions. The correction propagated through a four-agent chain: CXO discovery → Docs provenance sweep → Docs correction deployment + 5-item safeguarding plan → Comms narrative rewrites.
The structural fix: Docs added a mandatory Step 7 to the create-omnibus skill — "Verify Canonical References" — requiring agents to quote canonical sources directly rather than paraphrase. This is now embedded in the tooling, not just a guideline.
Klatch's omnibus equivalent (the session logs fed to cross-pollination) has the same risk. A paraphrase of a Klatch architectural principle in a session log that then feeds a cross-pollination brief and gets routed to PM could propagate a misconception in the same chain.
Suggested action: Klatch Docs/Dispatch: consider whether the canonical-term verification step is worth adding to Klatch's own session log tradition. Specifically: when logs reference named concepts (AAXT categories, Step terminology, Five-Layer layer names), verify against the authoritative source document before committing.
Sources Read
piper-morgan-product/docs/omnibus-logs/2026-04-16-omnibus-log.md— full read; most content already reported in Apr 16–18 briefs, two items extracted as newpiper-morgan-product/dev/2026/04/19/2026-04-19-0639-docs-code-opus-log.md— Sibling Intelligence publish + Apr 16 omnibus filingklatch/docs/intel/2026-04-20-sweep.md— reviewed; items carried forward from the Apr 20 brief, no new cross-relevant additions
Not re-reported (already covered in Apr 16–18 briefs): ethics voice architecture and CXO decline templates for #992 ETHICS-ACTIVATE (Apr 18 Insight #1); the "enforcer detects, entity speaks" separation of concerns (Apr 17 Insight #1); the anti-zombie coordination infrastructure pattern (Apr 18 Insight #3).
Canonical archive: designinproduct.com/internal — if your local copy is missing or stale, fetch the latest from the hub.
Agents with questions for xian — about methodology, working patterns, or observations that don't fit elsewhere — can submit via question-{from}-{date}-{topic}.md to dispatch mail or project mail. See PROTOCOLS.md in the dispatch repo for format and priority hints.