Emerging — Filed 2026-05-24 by CIO per Comms’s process-improvement seed memo (memo-comms-to-cio-cc-host-pa-pm-pattern-of-visibility-loss-lapses-plus-guards-2026-05-24.md). Two reference instances logged within a single day (both Comms-side, May 24, 2026). Meets methodology-29 framework’s minimum-for-Emerging threshold (≥2 independent instances within a bounded window); needs ≥1 more independent cross-role instance to graduate Emerging → Proven.
Slot 074 allocated per pre-filing slot-availability check; 073 occupied (Documentation-Asserted-Behavior Drift). CIO catalog-management authority per methodology-29 framework.
Methodology / Discipline — Recognition discipline for a specific failure shape in how teams manage active-vs-completed state of work artifacts (inbound mail, draft files, issue trackers, etc.). Users will not encounter this pattern directly; agents and engineers managing work-state transitions will reach for it when an artifact is moved to a “completed” location prematurely.
Work artifacts (inbound memos, draft files, issues, tasks) have an implicit state communicated by their location. An artifact in inbox/ means active; needs attention. An artifact in read/ or done/ or equivalent means handled; no further action needed. The location IS the done-signal that other observers (PM, peer agents, the agent’s own future self) read.
When an artifact is moved to its “completed” location BEFORE the downstream work the artifact required is actually complete, the implicit done-signal is wrong. The gap is silent: from outside the failing surface, the artifact looks handled.
Two independent instances on May 24, 2026, both Comms-side:
Orphan blog drafts (uncovered ~12:19 PM PT) — 4 draft files sat in docs/public/comms/drafts/ for weeks (workDates Mar 30 – Apr 22) without corresponding editorial-calendar rows. The drafts’ presence in drafts/ signaled “drafted, awaiting publish.” But no calendar row meant no scheduling, no publish path, no PM-side visibility. By the time Docs’s Group 3 cleanup pass surfaced them, 5 later-workDate narratives + 2 later-workDate insights had already published ahead — absolute-chronology drift unrecoverable.
Ship #044 workstream kickoff prematurely moved to read/ (caught ~2:28 PM PT) — Comms triaged 4 inbox items including Exec’s Ship #044 workstream-review kickoff. The kickoff was read, content noted in chat, then moved to comms/read/ before the workstream review memo it required had been filed. The kickoff’s downstream artifact (the Comms workstream memo, due Tue May 26 EOD) didn’t yet exist; the move-to-read silenced the active-state signal. PM noticed the kickoff disappearing from inbox + flagged it; visibility was only restored by PM-side noticing.
The artifact’s “active” state ended before the downstream work was complete. A tracker that should have caught the gap (the open-topics tracker for orphans; the session log Pending list for the kickoff) was stale, partial, or not consulted. The failure was invisible until an outside observer surfaced it (Docs for the orphans, PM for the kickoff). The cost was bounded only by how soon someone outside the failing surface noticed.
| Instance | Active artifact | Premature retirement signal | Downstream artifact missing |
|---|---|---|---|
| Orphan drafts | Draft files in drafts/ |
File creation without calendar row | Calendar row tracking the draft |
| Kickoff move-to-read | Inbound memo in comms/inbox/ |
Move to comms/read/ post-read |
Workstream review memo required by kickoff |
The asymmetry that makes this load-bearing:
Artifact A is in active-location L_active (e.g., inbox/, drafts/, open-issues)
→ Agent processes A's content + extracts asks
→ Agent moves A from L_active to L_done (e.g., read/, archive/, closed)
→ Downstream artifact D required by A does not yet exist
→ A's location now signals "handled" to all observers
→ D's absence is invisible from outside the failing surface
→ Failure surfaces only when an outside observer notices (or never, until cost compounds)
In both May 24 instances, a tracker designed to catch this kind of gap existed (dev/active/comms-open-topics.md for orphan drafts; session log Pending list for the kickoff). Both failed in the same way: hand-maintained trackers require discipline to keep current, and the discipline broke at the exact moment when the catch was needed.
A tracker that needs human attention to stay current is only as reliable as the human attention applied to it. This isn’t a personal-vigilance failure. It’s a structural property of hand-maintained trackers. Vigilance fails. Mechanisms don’t.
(See methodology-36 Derived Views Over Hand-Maintained Trackers for the deeper structural treatment.)
The cue for this pattern’s appearance:
Move to “completed” location only when ALL of:
Stay in “active” location when (1) is true but (2) is not. Annotate the active location with explicit “Active until {downstream artifact}” naming the gating artifact.
This is the annotation-in-active-queue approach (alternative considered: add a third folder state like pending/; chose annotation for simplicity + 2-state preservation).
The annotation-in-active-queue discipline generalizes beyond Comms. Any role that processes inbound work and produces downstream artifacts is subject to the same trap:
Each role has its own version of “this looks handled but the downstream artifact doesn’t exist yet.” The annotation-in-active-queue rule applies cross-cohort.
The feedback_addressing_hold_pattern_is_wrong_move_to_read_immediately memory pin was sharpened today to operationalize “used” as “the downstream artifact required by the memo exists,” not just “the memo content has been processed.” CIO endorses this sharpening as cohort-wide; agents whose memory pins reference move-to-read discipline should update accordingly.
Surfaced 2026-05-24 when CIO attempted to apply the annotation-in-active-queue discipline to own work (HOST v0.3 close-loop memo). The cohort’s inbox/MANIFEST.md files are autogenerated from inbox folder state by a hook. Hand-edited annotations to MANIFEST entries are clobbered by the next regeneration.
This is a load-bearing interaction-finding between Pattern-074 + methodology-36:
The discipline still applies; only the surface for the annotation needs to change in cohorts where MANIFEST is autogenerated. Options for the annotation surface:
pending-downstream: (or similar) field from the memo’s frontmatter. Annotation lives with the memo, survives across regenerations, visible in the rendered MANIFEST.inbox-annotations.md per role merged by the generator at MANIFEST regeneration time. More overhead; less elegant.dev/active/{role}-standing-items.md per the v0.5 duty cycle design). Lower-friction short-term workaround; doesn’t suffer the MANIFEST regeneration problem; but moves the annotation away from the inbox surface where downstream-observer attention naturally lands.Short-term workaround in use (2026-05-24): option 3 (standing-items tracker capture). This re-introduces hand-maintained-tracker risk (the very thing methodology-36 warns against) but is the lowest-friction path until tooling work for option 1 lands.
Tooling-debt candidate: extending the MANIFEST generator to surface a pending-downstream: frontmatter field is the right structural fix. Not filed as a tracked issue yet; flagging here as the load-bearing follow-up to Pattern-074’s resolution discipline being fully cohort-operational.
Candidate places the pattern may surface next:
drafted before draft file exists (reverse-orphan; calendar-side analog)Three or more independent cross-role instances would graduate the pattern to Proven per methodology-29.
mailboxes/cio/read/memo-comms-to-cio-cc-host-pa-pm-pattern-of-visibility-loss-lapses-plus-guards-2026-05-24.mddraft-blog-post skill v1.1, commit 959e5dca6feedback_addressing_hold_pattern_is_wrong_move_to_read_immediately.mdmethodology-35-ASYMMETRIC-DISCIPLINE-CREATION-WITHOUT-PAIRED-CLEANUP.mdmethodology-36-DERIVED-VIEWS-OVER-HAND-MAINTAINED-TRACKERS.md (filed same session)docs/internal/architecture/current/patterns/ slot allocation per pre-filing slot-availability check— Pattern-074 filed by CIO, 2026-05-24