206bf718bd
Grade one coverage line's five documentation items on two separated axes the old code conflated: required_state (from plan_type + payer_rules.json) and input_state (SUPPLIED | ABSENT). A line is "Clear to Ship" only when EVERY required item is satisfied by positive evidence — a required item with no data is NOT_EVALUATED and blocks green, never granted by absence of contradiction. - plan_type never guessed: None / unmapped / casing-variant -> "Plan Type Needed", structurally cannot be green; PECOS/PA sit NOT_EVALUATED while universal items (SWO/visit/diagnosis per LCD L33822) still grade. - Required-ness reads from payer_rules.json (single source of truth), removing the latent Python-set-vs-JSON divergence. - Config-surface fail-safe: a known plan missing/partial in config cannot grade green (NOT_EVALUATED), and the green gate checks every item. - Diagnosis accepts a real ICD-10 code as on-file; only explicit negatives gap. Additive only: not wired into the live pipeline. 27 tests green (full suite 73), code-auditor reviewed (H1/H2/M1/M2/M3/M4/L1 fixed). PROVISIONAL, flagged for Pi/Kisa: gap->label mapping (At Risk/Action Needed/ On Track) and the visit-recency scope boundary (timing layer, not readiness). Spec: docs/readiness-model-brief-2026-06-23.md section 4 Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
7.9 KiB
7.9 KiB
Signal — Ship-Status Ledger
Canonical, git-tracked record of what shipped, what didn't, and the reason or dependency for anything outstanding. Both Pi and Claude read and maintain this; Kisa validates. Ship-status never lives only in private memory. Update this on every ship or blocker change. Pairs with context/current-state.md (this = the detailed ledger; current-state = the live ACTIVE/NEXT snapshot).
Status legend: SHIPPED · IN PROGRESS · BLOCKED · NOT STARTED · DEFERRED · PI-SIDE · PARKED
Ledger (as of 2026-06-24)
| # | Item | Owner | Status | Reason / Dependency | Next action |
|---|---|---|---|---|---|
| 1 | Claude-side delegation trigger (dispatch-trigger.sh + settings.json UserPromptSubmit hook + dispatch thresholds) |
Claude | SHIPPED (verified live) | none | done |
| 2 | Audit Finding 1 — hook stdout injection works | Claude | SHIPPED | verified live (marker reached the model) | done |
| 3 | Audit Finding 2 — $1 bug → read stdin JSON |
Claude | SHIPPED | caught + fixed during build | done |
| 4 | Audit Finding 4 — old+new dispatch rules conflict | Claude | SHIPPED | added clarifying line (kept judgment fallback, marker = evaluate not auto-dispatch) | done |
| 5 | Convex-Signal spike brief | Claude | SHIPPED (committed this wrap) | — | Pi designs from it |
| 6 | Convex-Signal spike design | Pi | SHIPPED | docs/convex-signal-spike-design-2026-06-24.md. 40h/3-phase plan; Module map + Repository interface + Convex schema + Auth seam (Convex default, Clerk swappable) + Phase 1 (core readiness loop) = what Claude builds first. | Claude builds Phase 1 from it |
| 7 | signal-convex repo creation |
Claude (GitHub) + Kisa/token (Forgejo) | PENDING | Dependency: Pi design + Kisa approval (create once, correctly). Claude can do GitHub via gh; Forgejo has no API token here. |
after design approved |
| 8 | Convex BAA email (plan/cost/signing) | Kisa | IN PROGRESS | Kisa handling | Kisa sends |
| 9 | Readiness-model build (rename coverage→readiness, rules-to-config, grouping+verdict, device override, require-plan-type, frontend nesting) | Claude | IN PROGRESS | Step 3 SHIPPED 2026-06-26/27 (additive, not yet wired): (a) core/dedup.py grouping (CSVRow→MergedShipment→CoverageLine), 30 tests, auditor-reviewed; (b) core/readiness.py verdict engine — required-vs-input two-axis grading that KILLS the false green (required+absent never satisfied; clear-to-ship only on positive evidence; plan unknown→"Plan Type Needed"; required-ness read from payer_rules.json not Python sets), 27 tests, auditor-reviewed (H1/H2 config-surface false greens + M2/M3/M4 fixed). Added additive plan_type to ShipmentRecord. OPEN for Kisa/Pi: provisional gap→label mapping (At Risk/Action Needed/On Track) + H3 scope boundary (visit recency = timing layer, not readiness verdict — confirm). Remaining: rules-to-addressable-config (citation IDs), patient rollup, device-keyed override+recompute+rollup, require-plan-type enforcement + remove name-guesser/default in the LIVE path, frontend nesting + "plan type needed" UX (→ Pi). |
Next: hand frontend/label UX to Pi (brief staged); build patient rollup + WIRE orchestration in api/main.py/new pipeline module, NOT inside coverage_calculator (circular import). Each step auditor-reviewed. |
| 10 | NocoDB leads / CRM cleanup | Claude + Kisa | BLOCKED | Reason: leads table has no primary key (Signal-base data source bdmhi02nks1pg6g is corrupted — a fresh table there also came up PK-less; content base is healthy). SQLite can't add a PK to a populated table (Route 1 failed). Interim: capture still works (inserts via n8n signal-cgm-lead); the 2 test rows can't be deleted. |
resolved by the data-platform decision (item 11) |
| 11 | Data-platform decision (leads + future PHI backend) | Kisa (with Claude/Pi analysis) | OPEN | Options analyzed: repaired-NocoDB / Supabase Postgres / Neon / AWS (already set up, BAA+RDS+S3 parked) / Convex ($25/mo incl. BAA). Leads = non-PHI (stay outside PHI boundary). Decision informed by the Convex spike. | run spike → decide |
| 12 | Research-versioning Layer-1 (research-autopush.sh + Stop-hook entry) |
Claude | NOT STARTED | Dependency: Pi design exists (docs/research-versioning-design-2026-06-24.md). Lower priority than product. |
next infra slot |
| 13 | signal/docs/ auto-versioning (Audit Finding 5) |
Claude + Pi | NOT BUILT | Reason: research-versioning scope excludes signal/docs/. This is why design docs need manual commits. |
add signal/docs/ to scope |
| 14 | Pi research-versioning ext tool_call→tool_result (Finding 3) |
Pi | LIKELY DONE | Pi state reports persist gap closed (8 file writes this session); confirm it is the tool_result fix | Pi confirm |
| 15 | Mid-session commit push on clean tree (Finding 6) | Claude | DEFERRED | optional, low risk, adds session-close latency | revisit if loss occurs |
| 16 | Chief-of-staff sequential-dependency capability (Finding 7) | Claude | NOT TESTED | Dependency: a discovery test with a sequential multi-specialist brief | run the test |
| 17 | Pi injected-message behavior (Finding 8) | Pi | NOT TESTED | Pi-side test | Pi test |
| 18 | Skills-as-git-repo (Finding 9) | Claude/Pi | DEFERRED | Phase 2 per workflow-consolidation plan | phase 2 |
| 19 | Website copy + agent-file fixes | lead-content / Claude | STAGED, not deployed | prose rewrites routed to lead-content; BAA wording attorney-gated | deploy when ready |
| 20 | Signal AIOS product | — | PARKED | Activate only on demand-pull + Signal self-sustaining + delegable capacity | hold |
Timeline / critical path
- Next session: (a) Kisa gives the dedup decision → Claude builds the readiness model (item 9). (b) In parallel, Pi designs the Convex-Signal spike from the committed brief (item 6).
- After Pi's Convex design: Kisa validates → create
signal-convex(item 7) → Claude builds the spike. - Convex spike + BAA terms (item 8) → inform the data-platform decision (item 11) → which resolves the NocoDB leads/CRM fix (item 10).
- Infra cleanup (batched in slack): research-versioning Layer-1 +
signal/docs/scope (items 12–13, the reason docs need manual commits), then low-priority Findings 6/7/9. Findings 3/8 are Pi-side. - Fixed date: labABLE
/pilot-prepFri 2026-06-26 1:30 PM.
Dependencies summary
- Readiness build ⟸ Kisa dedup decision.
- Convex spike build ⟸ Pi design ⟸ committed brief (ready) ; and ⟸ Kisa approval + repo creation.
- NocoDB/leads fix ⟸ data-platform decision ⟸ Convex spike evidence + BAA terms.
- Design-doc durability ⟸
signal/docs/auto-versioning (until built, commit docs manually).
Maintenance
Update this whenever an item ships or a blocker changes. Move shipped items to a ## Shipped (archive) section when the table grows. Both agents reason from this; Kisa validates and redirects.
Design Verification (2026-06-24, Claude)
Pi's dedup design (item 9) and Convex-spike design (item 6) were verified line-by-line. Both are build-ready; fixes applied directly to the design docs:
- dedup-design-2026-06-24.md: (1) added a separate client-mapped
plan_typefield, distinct from displaypayer, so CoverageLine keys on the mapped plan type, never the payer name — the original draft reintroduced the killed plan-type-guessing bug. (2) Fixedmerge_quantitiesto sum across distinct component configs; the old blanket-max dropped quantity in mixed duplicate+split groups (e.g. two [sensor] + one [transmitter] returned 3 instead of 4). - convex-signal-spike-design-2026-06-24.md:
shipments.orderNumber/hcpcschanged to arrays (post-dedup multi-value); §11 dedup note de-staled to LOCKED; carry-forward note to grade on the mappedplanType. - Noted, not a blocker: "Convex auth -> Clerk = config-only swap" is optimistic; treat as low-but-nonzero effort (Convex Auth still maturing).