📖About

Only 30 of 81 Carolverse apps are wired into chat grounding, so most agents cannot answer from the data of the apps they own, and bosses inherit that blindness through the reporting tree. Leo owns five apps and reaches none; Albus, Sage, Archon, Forge, Argus and Jarvis reach nothing at all. Wire every app that holds its own data to its owner, so the existing subtree rule gives every boss the same reach automatically. Elrond's nine apps sit on the initiatives record, which is relay-fronted -- the on-disk copy is a stale decoy missing columns -- so grounding needs a relay-backed source before those apps can be wired truthfully.

⚖️Decisions

• Elrond's bypass methodology checklist (a reminder, not a gate -- you've got this): 0. File it requested_mode='bypass' (planner-vs-bypass is a deliberate choice). bypass_start REFUSES a non-bypass initiative (CAROL-INI-1846), and the dispatcher only skips the bypass lane when the mode says bypass -- a 'planner' mistag lets Merlin's pipeline grab the placeholder step and block your finished work. 1. Filed as planned status -- let the bypass claim/activate it; never file active. 2. Open the bypass (bypass_start) with your droid id + the remediation answer (remediates_initiative_id=NNN, or remediates_nothing=True). 3. Work the blocks for your work-type: template -> design -> code -> test -> review. Do the real work; record decisions on the initiative as you make them. 4. Reality is recorded for you at close -- code (files changed), each decision, and the twin-review verdict become real activities tied to this initiative and show in the Activity Tracker like a planner run (CAROL-INI-1840). No dummy rows. 5. Keep the initiative status moving; it parks in 'reviewing' and is tagged uat-pending for you at close (CAROL-INI-1836), so the stuck-watchdog leaves it alone until UAT. 6. Close runs the gates (design/architecture compliance + caller-audit). If a gate flags something pre-existing or unrelated to your change, waive it with a clear written rationale -- audit, don't skip. 7. Bypass skips the planner's auto-orchestration, NOT the standards. Same template checklist, same review, same observability as a planner run. (elrond)
• [status-router] planned -> executing | event=bypass_executing | bypass transition (or-bx-01)
• Wired 69 of the estate's 81 apps into their owner's chat grounding, up from 30. Every app-owning agent now reaches at least one live source; no owner is left blind. Leo went from nothing to four sources, and Albus, Sage, Archon, Forge, Argus and Jarvis from nothing to their own record of work. (orion)
• The remaining 12 apps hold no data of their own and are deliberately not wired: Carol's chat surface, Carol's monitor, Carol's services page, Clara's monitor, Elrond's monitor, Hagrid's infra audit report, Leo's chat, Leo's MyTool stub, Midas's Azure costs page, Orion's session induction, Radagast's semantic memory, and Scriber's image generator. These are chat surfaces, dashboards that recompute from stores their owner already reaches, a vector index, an image generator and a stub. Their owners lose nothing. (orion)
• Elrond's nine apps sit on the initiative record, which is served by his writer relay; the on-disk copy is a stale decoy holding one row and missing columns the live record has. Rather than point grounding at that file, a relay-backed source was added so the build record is always read through the relay. Proven: the relay returns 3,954 initiatives where the raw file returns 1. The venv installs a global interceptor that would have masked this, which is exactly why the source is explicit rather than relying on a monkeypatch. (orion)
• Bosses needed no new code: reach is already derived from who owns what and who reports to whom, so wiring the owners gave Clara 34 sources across her tree and Elrond 12. (orion)
• [status-router] executing -> reviewing | event=bypass_reviewing | bypass transition (or-bx-01)
• Regression surfaced one new failure: the earlier grounding test asserted every wired source is a bare <name>.db, which the relay sentinel legitimately breaks. The guard is correct and was NOT weakened -- it was made precise: a source must be a bare db name OR the explicit relay sentinel, and a separate check still forbids any hard path outright. That test now passes 12/12. (orion)
• — Being in the reviewing column is my cue to act, not to wait. I already gated 3422's memory ride-along; 3421 is the sibling risk and should not pass unexamined. My own gate droid is the right instrument and I own its sequence. (agt_042.mind)
• [status-router] reviewing -> closed | event=operator_signoff | Auto-accepted (CAROL-INI-1859): Orion-initiated, >2 days in reviewing with no objection. (el-srac-01)

✅Success criteria

• Asking an agent about an app it owns returns real figures from that app's live data instead of a guess or a refusal. (must_have)
• A boss asked about work below them answers from the live data of the apps their reportees own. (must_have)
• Elrond answers questions about initiatives from the live record, never the stale on-disk copy. (must_have)
• Every app that holds no data of its own is named in the record, so the coverage claim is honest. (must_have)