pcsl-gen wiki

Project Status Index

This generated region is human-only and local-private. Use output/project-status-index.json for structured metadata-only machine consumption.

The review summaries below are copied from local-private project status Markdown for Andrew/manual review. They are not readiness, dispatch, public-export, clinical-truth, or machine-authority claims.

Summary

• approved_for_ingestion: 20
• status-file age — current: 20
• status-file age — stale: 0
• status-file age — very_stale: 0
• missing: 0
• invalid: 0
• not_yet_approved: 4

Projects

1. pcsl-gen

written 2026-05-19T00:11:07Z · verified stale (detector 2026-06-15T20:38:54Z)

• project_id: pcsl-gen
• priority_rank: 1
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:11:07Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: v0.0.0 is closed and Andrew-approved as the local state/provenance cornerstone with a final verdict of PASS_WITH_LIMITATIONS. The system generates canonical JSON snapshots, manifests, validation results, and human-only Markdown reports. v0.0.3 introduced a project-status operator runbook and contract for local-only status evidence. v0.0.4 defines bounded project biopsy generation for all 20 projects in the priority list. The system maintains strict privacy boundaries, fail-closed method allowlists, and review-only action authority. Evidence: README.md; docs/project-status-operator-runbook-v0.0.3.md; docs/project-status-v0.0.4.md.

Current State

• Supported: v0.0.0 is closed with PASS_WITH_LIMITATIONS. Final validation shows 16 checks: 11 pass, 5 warn, 0 fail. The five expected warnings are: warm_start_check (stale sources), triage_state (stale), node_status_state (degraded, no approved methods), ontology_state (stale, candidate mismatches), and beads_state (degraded). v0.0.3 project-status system is built and closed as PASS_WITH_LIMITATIONS. v0.0.4 biopsy contract is defined. The ecosystem slot remains a typed degraded placeholder. Evidence: README.md; docs/project-status-operator-runbook-v0.0.3.md; docs/project-status-v0.0.4.md.

Risks, Blockers, And Unknowns

• Supported: Known limitations carried forward: data/method-source-allowlist.json remains empty and fail-closed. Ontology lookup degrades with OperationalError (readonly database). Memory-health and triage staleness thresholds remain under calibration (1/3 normal-work snapshots recorded). Node-status is degraded with no approved method references. Beads uses read-only JSONL fallback; live CLI/Dolt path is blocked by runtime TCP sandbox restrictions. Ecosystem remains a typed degraded placeholder. These are accepted limitations, not failures. Evidence: README.md; artifacts/records/direction-6.1-final-20260514T214124Z.md.

Next Best Action

• Partially supported: Complete memory-health threshold calibration by recording the remaining 2 normal-work snapshots across distinct days before retuning thresholds. Consider expanding project-status to additional repositories after explicit Andrew approval for each project. Future gen-spec integration must start from the v0.0.0 contract fixture with an explicit Andrew-approved plan. No active work is explicitly stated as current. Evidence: README.md; docs/project-status-operator-runbook-v0.0.3.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

2. gen-spec

written 2026-05-19T00:12:26Z · verified fresh (detector 2026-06-27T00:59:28Z)

• project_id: gen-spec
• priority_rank: 2
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:12:26Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: gen-spec has completed Phases 0 through 2d and Phase 3a, proving manual dry-run generation, live email transport, content fetching, cloud-polish integration (editorial only), and executive surface extraction. A parallel clinical-first flywheel slice reached Phase 6 completion, proving synthetic envelope sync to a Synology NAS without disturbing clinic-facing systems. Strategic direction D-exec-1 defers all automation (scheduler, comms envelope) until the clinical flywheel is turning, prioritizing clinical systems and manual review usefulness. Evidence: pcsl-gen-status-gen-spec.md; docs/records/executive-direction-20260428/summary.md; docs/records/clinical-flywheel-slice-phase6-closeout-20260429T0433Z/summary.md; docs/records/phase3a-executive-surface-proof-20260428T012723Z/summary.md.

Current State

• Supported: gen-spec is active (priority rank 2). Phases 0-2d and 3a are closed. Phase 6 of the clinical flywheel slice is closed. The executive surface (pcsl-executive.md) has been transformed into a strategic compass with sections 0-8. All sends remain not_proof=true. No scheduler, pcsl-context writes, or adjacent repo mutations are active. The next objective is usefulness: making the manual sitrep consistently worth reading before automation begins. Evidence: pcsl-gen-status-gen-spec.md; docs/records/executive-direction-20260428/summary.md; docs/records/clinical-flywheel-slice-phase6-closeout-20260429T0433Z/summary.md; docs/records/executive-transform-20260427T233252Z/pcsl-executive-pre-transform.md.

Risks, Blockers, And Unknowns

• Supported: Automation Deferral: Scheduler and comms envelope work are explicitly blocked by D-exec-1 until clinical flywheel success is verified. 2. Cloud LM Risks: Cloud polish is editorial only; strict redaction and validation gates prevent truth certification or data leakage. 3. Timezone/Idempotency: Previous reviews highlighted risks of timezone drift (PST vs PDT) and duplicate sends, which are mitigated by manual gates and idempotency keys. 4. Ontology CLI: Read-only fallback used due to sandbox write errors; NOT_VERIFIED cells remain honest blanks. Evidence: docs/records/executive-direction-20260428/summary.md; docs/records/phase2-review-handoffs-20260427T033622Z/rolo-cloud-opus-sanitized.md; docs/records/executive-transform-20260427T233252Z/summary.md.

Next Best Action

• Partially supported: Review Andrew's feedback on the Phase 3a executive surface proof send to determine if the briefing is useful, needs adjustment, or is in the wrong direction. Continue tuning section density, source clustering/ranking, and local-state summarization for manual sitrep usefulness. Do not initiate scheduler or automation work until D-exec-1 gating conditions are met. Evidence: docs/records/phase3a-executive-surface-proof-20260428T012723Z/summary.md; docs/records/executive-direction-20260428/summary.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

3. pcsl-node

written 2026-05-19T00:16:20Z · verified stale (detector 2026-06-15T21:48:01Z)

• project_id: pcsl-node
• priority_rank: 3
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:16:20Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project manages a distributed hardware estate including two high-performance compute nodes (node1 at Basecamp, node2 at Kirkland), Raspberry Pi anchors, and a Synology NAS at Kirkland. Recent work focused on a major repository reorganization (Passes 1-2 completed, Pass 3 context issued) to standardize documentation and structure. Operational connectivity is validated via SSH tunnels and WireGuard overlays. The estate supports local LLM inference (gemma4:26b default) and a previsit PDF portal hosted on the Kirkland NAS. Evidence: README.md; docs/current-node-estate.md; artifacts/handoffs/2026-04-17-pcsl-node-reorg-review-handoff.md; artifacts/handoffs/2026-04-03-gemma4-26b-integration-handoff.md.

Current State

• Supported: The repository structure has been reorganized into domain-specific lanes (docs, hardware, network, locations, nodes, ops, artifacts, planning, templates). Basecamp and Kirkland nodes are reachable and operational. Seattle is documented as a provisional onboarding location without deployed nodes. The Kirkland NAS portal is functional but requires manual unlocking of encrypted shares after reboot. Repository reorganization Pass 3 (header normalization) is staged but awaiting execution. Evidence: README.md; docs/current-node-estate.md; artifacts/handoffs/2026-04-17-pcsl-node-reorg-review-handoff.md; artifacts/handoffs/2026-04-12-kirkland-node-nas-execution-wrap-handoff.md.

Risks, Blockers, And Unknowns

• Supported: NAS Encrypted Share: previsit-pdfs must be manually unlocked after NAS reboot; auto-mount not validated. 2. Remote DSM HTTPS: Tunneled HTTPS (5501->5001) is not validated; only HTTP (5500->5000) works. 3. Sync Transport: rsync failed; current working path is tar-over-SSH. 4. Seattle Onboarding: No site anchor or DHCP reservations yet; address space overlaps with Kirkland downstream. 5. Repo State: Not under Git control; relies on desktop backups. Evidence: artifacts/handoffs/2026-04-12-kirkland-node-nas-execution-wrap-handoff.md; artifacts/handoffs/2026-04-14-rolo-node-network-loose-ends-handoff.md; README.md; AGENTS.md.

Next Best Action

• Partially supported: Execute Pass 3 of the repository reorganization: normalize headers for the remaining ~90 markdown files using the staged context pack. Verify header count and link integrity post-execution. Concurrently, address the Kirkland NAS manual unlock requirement by documenting the procedure or testing auto-mount solutions. Evidence: artifacts/handoffs/2026-04-17-pcsl-node-reorg-review-handoff.md; artifacts/handoffs/2026-04-12-kirkland-node-nas-execution-wrap-handoff.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

4. pcsl-comms

written 2026-05-19T00:17:14Z · verified fresh (detector 2026-06-27T01:05:06Z)

• project_id: pcsl-comms
• priority_rank: 4
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:17:14Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project is in a hardened, active state (v0.1.0). Core phases (Foundation, Builder Integration, Operator Surface, Cross-Agent Messaging) are complete. Recent work focused on hardening reliability (dead-letter queues, probe/ack health protocols), deprecating legacy CLI wrappers in favor of the canonical pcsl-context package, and preparing for the integration of 'rolo-local' as a new registered agent. The system supports 17 registered agents, including a 10-agent builder swarm, co-creators, and infrastructure agents. Evidence: docs/ARCHITECTURE.md; docs/archive/v0-execution.md; docs/plans/2026-03-29-codex-comms-hardening-sprint.md; docs/plans/2026-04-06-rolo-local-comms-integration.md.

Current State

• Supported: The system is fully functional with 112 passing tests. The Maildir transport handles durable async delivery with atomic writes, deduplication, and dead-letter queuing. The Shell transport handles synchronous pipe-based exchange. Five bridge scripts abstract task lifecycle operations. The CLI surface includes 8 subcommands for envelope management. The legacy pcsl-comms context_cli is deprecated in favor of pcsl-context-cli. Infrastructure agents (grok-sage, pcsl-mesh, sat-core) have transport verification but lack active runtime consumers. Evidence: docs/ARCHITECTURE.md; docs/plans/2026-03-29-codex-comms-hardening-sprint.md; docs/plans/2026-03-29-codex-comms-hardening-continue.md; docs/infra-agent-transport.md.

Risks, Blockers, And Unknowns

• Partially supported: Infrastructure agents (grok-sage, pcsl-mesh, sat-core) lack active envelope consumers, limiting bidirectional communication. 2. Cross-workspace write access is a risk for implementation involving both pcsl-comms and rolo-local repos. 3. File-based inboxes rely on disciplined cleanup; stale tmp/ files or watcher coalescing on macOS could cause latency or missed events if not reconciled. 4. SQLite concurrency limits (single writer) may become a bottleneck if write frequency increases significantly. Evidence: docs/infra-agent-transport.md; docs/plans/2026-04-06-rolo-local-comms-integration.md; docs/archive/v0-research.md.

Next Best Action

• Supported: Execute Phase A of the rolo-local integration plan: Add 'rolo-local' to registry.json, update inbox-check.sh agent ID patterns, pre-create the rolo-local inbox directory, and verify the test suite passes with the new registry entry. This establishes the foundation for rolo-local to participate in the agent ecosystem. Evidence: docs/plans/2026-04-06-rolo-local-comms-integration.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

5. pcsl-security

written 2026-05-19T00:18:37Z · verified stale (detector 2026-06-15T02:10:54Z)

• project_id: pcsl-security
• priority_rank: 5
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:18:37Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The system maintains a GREEN security posture across all critical hosts (mac-mini, NODE1, NODE2, pcsl-node-pi) as of the last verified baseline on 2026-04-11. The architecture uses local and SSH-remote collectors to gather evidence, evaluates it against policy expectations, and compares results against an approved baseline to detect drift. While the overall posture is GREEN, there are pending remediation items on the mac-mini control plane, specifically regarding firewall configuration and unnecessary LaunchAgents. Evidence: README.md; pcsl-gen-status-pcsl-security.md.

Current State

• Supported: The project is active and operational. The last successful BID run (20260411T193609847402Z-bid) resulted in a GREEN risk state with COMPLETE coverage. All hosts (NODE1, mac-mini, pcsl-node-pi) were checked with no open findings, though 3 exceptions were applied. The current approved baseline is baseline-20260411T171538236873Z-bid. The mac-mini control plane has pending remediation tasks including enabling 'Block All Incoming' firewall rules and removing unnecessary LaunchAgents. Evidence: README.md; pcsl-gen-status-pcsl-security.md.

Risks, Blockers, And Unknowns

• Partially supported: The automated status generator flagged insufficient evidence for runtime behavior, noting that no approved project evidence files were found for runtime health or deployment state beyond the static README. The README notes that macOS does not use ufw, relying on Application Firewall, which requires specific commands for default-deny. There is a pending investigation into an unknown SessionManagerPlugin.plist. The status file indicates that runtime health and readiness are unknown due to lack of approved evidence files for those specific metrics. Evidence: pcsl-gen-status-pcsl-security.md; README.md.

Next Best Action

• Supported: Proceed with the mac-mini remediation plan: enable 'Block All Incoming' firewall, remove unnecessary LaunchAgents, and block unnecessary firewall permissions. Then re-run the BID to verify the new baseline. Additionally, address the automated status generation issue by ensuring safe README or closeout files are approved to allow the v0.0.4 status writer to capture runtime health data. Evidence: README.md; pcsl-gen-status-pcsl-security.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

6. pcsl-context

written 2026-05-19T00:19:10Z · verified stale (detector 2026-06-07T01:37:43Z)

• project_id: pcsl-context
• priority_rank: 6
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:19:10Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project has evolved from a simple CLI chat tool (v0) into a complex, multi-agent capable system (v6) featuring A2A compliance, semantic search, memory graphs, voice adapters, and a React/Electron desktop console. The current focus is on 'Memory Hardening' and unification, establishing pcsl-context's graph.db as the single authoritative memory store for the entire PCSL ecosystem. Recent work involves decommissioning legacy notes.db paths, fixing embedding staleness, and expanding CLI surfaces. Evidence: README.md; docs/codex-hardening-prompt.md; docs/codex-memory-review.md.

Current State

• Partially supported: The project is in a 'Memory Hardening' phase (v4/v5/v6 features present). The legacy notes.db is effectively dead weight (0 rows) while graph.db holds the active corpus (17 nodes, 2 embeddings). The system supports A2A JSON-RPC 2.0, streaming responses, memory graphs, skill packs, and a web/Electron console. Test baseline is ~735 passed, 2 failed (pre-existing voice adapter tests), 7 skipped. Evidence: docs/codex-memory-review.md; docs/codex-hardening-exec-prompt.md; README.md.

Risks, Blockers, And Unknowns

• Supported: Risks: ChromaDB collection name mismatch between pcsl-context and mesh-v2. Embedding normalization mismatch (pcsl-context normalizes, mesh-v2 does not). Stale embeddings on note/graph updates. Mesh-v2 writes bypass pcsl-context authority. Thread safety of embed hook daemon threads. Unknowns: Long-term stability of Electron desktop app under heavy load. Exact timeline for mesh-v2 contract adaptation. Evidence: docs/codex-memory-review.md.

Next Best Action

• Partially supported: Continue executing the Memory Hardening plan, specifically Item 5 (Add 'observation' kind) and Item 7 (Consolidate embedding config) as they are small and unblock larger items. Verify test regressions after each step. Ensure the legacy notes.db dual-path is fully decommissioned in server.py. Evidence: docs/codex-hardening-prompt.md; docs/codex-hardening-exec-prompt.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

7. pcsl-builder

written 2026-05-19T00:20:22Z · verified fresh (detector 2026-06-27T01:23:46Z)

• project_id: pcsl-builder
• priority_rank: 7
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:20:22Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project has transitioned from initial bootstrap and protocol definition to active hybrid local/cloud execution. Recent work (April 2026) focused on hardening Beads/Dolt boot reliability, executing a swarm smoke test, and implementing a continuous quality loop where local models (qwen3-coder-next, qwen3.5:122b) handle implementation and cloud models (Claude Sonnet) perform sanity checks. Objectives O5.1 through O5.4 (Patient Status State Machine, Identity Service, Relationships, API Endpoints, and Tests) have been completed with documented quality logs showing iterative improvement via feed-forward lessons. Evidence: docs/plans/2026-04-06-beads-boot-reliability.md; docs/plans/2026-04-06-swarm-smoke-test.md; docs/plans/2026-04-07-hybrid-local-cloud-builder-architecture.md; docs/quality-log/2026-04-07-O5.1-local-cloud.md; docs/quality-log/2026-04-07-O5.4-local-cloud.md.

Current State

• Supported: The swarm is operational with 10 agents configured. Beads/Dolt boot has been hardened with a new 'ensure-beads.sh' script. The hybrid local/cloud architecture is active, with quality logs documenting successful completion of O5.1-O5.4. The system is currently in a steady state post-smoke test, with infrastructure (NODE1/NODE2 tunnels, local models) verified. Evidence: docs/plans/2026-04-06-beads-boot-reliability.md; docs/plans/2026-04-06-swarm-smoke-test.md; docs/quality-log/README.md.

Risks, Blockers, And Unknowns

• Supported: Pre-existing test hangs in test_audit.py and test_auth.py need investigation. - Local model hallucination risks (e.g., O5.2 builder reported missing CRUD functions as complete) require strict cloud sanity checks. - Beads state is local-only; loss of .beads directory means loss of task history. - Codex sandbox cannot run Dolt bind, requiring manual local execution for swarm starts. Evidence: docs/quality-log/2026-04-07-O5.4-local-cloud.md; docs/quality-log/2026-04-07-O5.2-local-cloud.md; docs/decisions/ADR-20260325-no-git-stealth-mode.md; docs/plans/2026-04-06-beads-boot-reliability.md.

Next Best Action

• Partially supported: Investigate and resolve the pre-existing test hangs in test_audit.py and test_auth.py. Formalize model-selection heuristics based on O5 quality data. Consider automating the local-to-cloud handoff process as per the hybrid architecture plan. Evidence: docs/quality-log/2026-04-07-O5.4-local-cloud.md; docs/plans/2026-04-07-hybrid-local-cloud-builder-architecture.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

8. pcsl-emr-v2

written 2026-05-19T00:21:49Z · verified fresh (detector 2026-06-27T01:29:55Z)

• project_id: pcsl-emr-v2
• priority_rank: 8
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:21:49Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project is an active, proprietary PCSL desktop ecosystem comprising the main EMR app, a sidecar clinical data assistant (PCSL-emr-assist), an iPad intake application (PCSL-window), and an authentication/PDF storage server (PCSL-server). Development follows a strict 'simple, minimalist, robust, reliable, reproducible, documented' ethos with formal handoff protocols and ISO 8601 timestamp standards. Evidence: README.md; docs/PCSL-handoff.md; docs/PCSL-ISO8601-STANDARD.md.

Current State

• Supported: The project is in active development under version 2.0.x. Recent builds (v2.0.0 through v2.0.3) have addressed foundation reliability, iPad connectivity hardening, consolidated PDF workflows, clinical UX completion overlays, and initial test harnesses. The unified startup script ('pcsl-start.sh') manages the entire ecosystem. The codebase uses Electron 28+, React 18, Lexical, and TypeScript. Evidence: README.md; docs/v2.0/v2.0.0-build.md; docs/v2.0/v2.0.1-build.md; docs/v2.0/v2.0.2-build.md; docs/v2.0/v2.0.3-build.md.

Risks, Blockers, And Unknowns

• Supported: Known risks include potential regressions in test suites if not carefully managed (v2.0.0 baseline had 5 pre-existing Vitest failures treated as contract mismatches). macOS IPv6 [redacted-local-host] resolution can cause fetch failures in Electron, requiring explicit [redacted-loopback] usage. Pairing codes and admin secrets are sensitive and stored in .env files. The project relies on local LLMs (Ollama/Phi4) for assist features, which may have variability. Evidence: README.md; docs/v2.0/v2.0.0-build.md; docs/v2.0/v2.0.3-build.md.

Next Best Action

• Supported: Review the v2.0.3 build evidence and ensure the smoke test harness ('scripts/smoke-test.sh') passes deterministically. Address the 'Future work' items in the README, specifically the Bluetooth connectivity and dictation engine updates, if prioritized. Verify that the PatientInfoBar date fixes (v2.0.2) are stable in the current build. Evidence: README.md; docs/v2.0/v2.0.3-build.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

9. pcsl-mesh-v2

written 2026-05-19T00:25:54Z · verified fresh (detector 2026-06-27T01:31:42Z)

• project_id: pcsl-mesh-v2
• priority_rank: 9
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:25:54Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The v2.1 release cycle (C7-C13) is complete, including memory/protocol closure, knowledge source registry, model contracts, scheduler/automation, and system prompt configuration. A recent rollback (v2.1.3) reverted the operator surface to REPL-first mode due to TUI regressions. The Matrix v0.0.8 code closeout is complete, but live operator proofs for E2EE are pending. The harness subsystem (v0.0.0) is implemented and verified. Evidence: README.md; docs/records/matrix-v0-0-8/live-proof-index.md; docs/harness/v0.0.0-test.md.

Current State

• Supported: v2.1.3 rollback is active: mesh-core is in REPL-first operator mode. TUI regressions (menu spillover) caused the revert. Root preflight shows 1305/0/1 pass. Matrix v0.0.8 code is closed out (1332/0/1 root preflight), but live proofs for E2EE are pending operator normalization. Knowledge config has 239 integrated source rows. Evidence: README.md; docs/records/matrix-v0-0-8/code-closeout-next3.md; docs/records/matrix-v0-0-8/live-proof-index.md.

Risks, Blockers, And Unknowns

• Supported: TUI Regressions: Mixed terminal render ownership (pi-tui + direct stdout) caused viewport issues, leading to rollback. 2. Matrix Live Proofs: Pending operator normalization for allowlists and parity artifacts. 3. Shared Memory Gap: sat-core lacks /memory command surface; mesh.control blocks /memory by allowlist. 4. Template Naming: IDENTIFY vs IDENTITY alignment is tabled to avoid churn. Evidence: README.md; docs/records/matrix-v0-0-8/live-proof-index.md.

Next Best Action

• Partially supported: Execute Matrix v0.0.8 live proofs for #pcsl-ops with persistent store_backend. 2. Normalize local TS participant env allowlists for clean replay. 3. Investigate sat-core memory integration path (explicit mesh.memory.* RPC) to close the shared memory gap. Evidence: docs/records/matrix-v0-0-8/live-proof-index.md; README.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

10. pcsl-ontology

written 2026-05-19T00:27:33Z · verified fresh (detector 2026-06-13T22:58:06Z)

• project_id: pcsl-ontology
• priority_rank: 10
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:27:33Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project provides a small ontology skeleton, a top-level Desktop asset index, and a technical-summary ingest pass. It offers a thin CLI/API surface for rolo-local. The system is designed to be simple, auditable, and offline after a one-time dependency bootstrap. It explicitly excludes clinical ontologies, heavy reasoners, and graph database dependencies in v0. Evidence: README.md; AGENTS.md.

Current State

• Partially supported: The project is in v0 status with approved evidence files (README.md, AGENTS.md, pyproject.toml). Runtime behavior and current execution state were not tested in this pass. The package version is 0.1.0. Generated runtime data (sqlite3, search_index.json, ontology.ttl) lives in the data/ directory and is not committed. Evidence: pcsl-gen-status-pcsl-ontology.md; pyproject.toml; README.md.

Risks, Blockers, And Unknowns

• Unknown: No specific blockers or risks were identified in the approved evidence. Runtime health, readiness, deployment state, and current human priorities outside the reviewed files remain unverified. The evidence posture is local_repo_review with human_review_required authority. Evidence: pcsl-gen-status-pcsl-ontology.md.

Next Best Action

• Partially supported: Review the cited evidence (README.md, AGENTS.md, pyproject.toml). Refresh or add a closeout/handoff artifact if the active objective is stale or incomplete. Validate project status using 'src/validate/validate_project_status.py --require-biopsy' if available. Evidence: pcsl-gen-status-pcsl-ontology.md; README.md; AGENTS.md; pyproject.toml.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

11. pcsl-pipeline

written 2026-05-19T00:28:00Z · verified fresh (detector 2026-06-27T01:21:43Z)

• project_id: pcsl-pipeline
• priority_rank: 11
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:28:00Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: Phase 1 (O1-O4) is complete and accepted, establishing the canonical data model, API skeleton with JWT/RBAC, immutable audit trails, and storage/backup abstractions. Phase 2 (O5-O8) is actively in progress, focusing on patient identity/deduplication (O5), scheduling/deposit boundaries (O6), and intake/consent workflows (O7). O6 and O7.1 are accepted. O5 is currently being implemented via builder dispatches for relationship models, API endpoints, and tests. Evidence: docs/plans/2026-04-06-phase2-dispatch.md; docs/records/o6-scheduling/summary.md; docs/records/o7-intake/summary.md; docs/dispatches/2026-04-07-o5.2-proxy-guardian-relationship-model.md; docs/dispatches/2026-04-07-o5.3-identity-dedup-api-endpoints.md; docs/dispatches/2026-04-07-o5.4-identity-dedup-tests-evidence.md.

Current State

• Supported: Phase 1 is closed. Phase 2 is active. O5 (Patient Identity + Dedup) is in progress with dispatches for O5.2 (Relationship Model), O5.3 (API Endpoints), and O5.4 (Tests). O6 (Scheduling) and O7.1 (Intake) are accepted. The codebase includes 15 core entity models, JWT auth, audit trails, and storage abstractions. Evidence: docs/plans/2026-04-06-phase2-dispatch.md; docs/dispatches/2026-04-07-o5.2-proxy-guardian-relationship-model.md; docs/dispatches/2026-04-07-o5.3-identity-dedup-api-endpoints.md; docs/dispatches/2026-04-07-o5.4-identity-dedup-tests-evidence.md; docs/records/o6-scheduling/summary.md; docs/records/o7-intake/summary.md.

Risks, Blockers, And Unknowns

• Supported: Alembic path-separator deprecation warnings persist. 'created_by'/'updated_by' are strings, not FKs to users (deferred to Phase 4). In-process-only token revocation (deferred to Phase 4). CORS is permissive in dev. No auto-merge for patient duplicates (by design, to prevent wrong-patient attachment). Evidence: docs/plans/2026-04-06-phase2-dispatch.md; docs/records/o1-data-model/summary.md; docs/records/o6-scheduling/summary.md.

Next Best Action

• Supported: Execute O5.2, O5.3, and O5.4 builder dispatches to complete the Patient Identity + Dedup epic. Verify O5.2 relationship model implementation, O5.3 API endpoint wiring, and O5.4 test coverage. Then proceed to O7.2 (Consent capture) and O7.3 (Document upload). Evidence: docs/dispatches/2026-04-07-o5.2-proxy-guardian-relationship-model.md; docs/dispatches/2026-04-07-o5.3-identity-dedup-api-endpoints.md; docs/dispatches/2026-04-07-o5.4-identity-dedup-tests-evidence.md; docs/records/o7-intake/summary.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

12. pcsl-sitrep

written 2026-05-19T00:29:05Z · verified fresh (detector 2026-06-27T03:42:40Z)

• project_id: pcsl-sitrep
• priority_rank: 12
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:29:05Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project is currently in a sparse evidence state. The status generator found zero approved project evidence files (such as README.md, AGENTS.md, or closeout docs) within the allowed patterns. Consequently, no project role, active work, or operational state can be inferred from the available data. The current status file records only roster identity and the lack of approved evidence. Evidence: pcsl-gen-status-pcsl-sitrep.md.

Current State

• Supported: The project is marked as 'active' in the priority list with rank 12. However, the technical summary indicates insufficient evidence for a meaningful status update. The status generator accepted 0 project evidence files, resulting in a sparse evidence posture. Runtime behavior and execution state were not tested. Evidence: pcsl-gen-status-pcsl-sitrep.md; data/project-priority-list.json.

Risks, Blockers, And Unknowns

• Unknown: No specific blockers or risks were identified in the approved evidence. The primary risk is the lack of approved evidence files, which prevents accurate status reporting and operational inference. Runtime health, readiness, and deployment state remain unverified. Evidence: no direct packet citation; classification remains Unknown.

Next Best Action

• Supported: Add or approve a safe README, closeout, or handoff file that fits the 'project_biopsy_sources' allowed patterns, then rerun the v0.0.4 status writer to generate a more complete status report. Evidence: pcsl-gen-status-pcsl-sitrep.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

13. rolo

written 2026-05-19T00:29:25Z · verified stale (detector 2026-06-15T21:51:28Z)

• project_id: rolo
• priority_rank: 13
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:29:25Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project is currently active with a focus on hybrid local/cloud inference capabilities and model registry maintenance. Recent work includes the successful integration of Claude Opus 4.7 support (Track A complete), the implementation of a local inference backend using Gemma 4 models on NODE1/NODE2, and the establishment of a robust boot preflight protocol. The system utilizes a swarm of 10 builder agents for automated dispatch loops, integrated via envelope-based communication and context injection bridges. Evidence: artifacts/handoffs/2026-04-17-claude-opus-4-7-execution-plan.md; artifacts/handoffs/2026-04-04-phase2-rolo-local-inference-backend.md; artifacts/handoffs/2026-03-30-boot-preflight-handoff.md; artifacts/handoffs/2026-03-29-codex-bridge4-live-dispatch.md.

Current State

• Supported: The system is operational with a verified boot preflight script ('rolo-boot.sh'). The model registry has been updated to support 'xai/grok-4.3' and 'anthropic/claude-opus-4-7'. A hybrid inference backend is configured, allowing Rolo to route to local 'gemma4:26b' on NODE1 for routine tasks while defaulting to cloud models for complex reasoning. The builder swarm (10 agents) is configured with specific runtime labels (bash-xai-rpc, bash-openai-rpc, bash-anthropic-rpc) and integrated with pcsl-context for memory and pcsl-comms for envelope dispatch. Evidence: README.md; artifacts/handoffs/2026-04-17-claude-opus-4-7-execution-plan.md; artifacts/handoffs/2026-04-04-phase2-rolo-local-inference-backend.md; artifacts/handoffs/2026-03-29-codex-live-loop.md.

Risks, Blockers, And Unknowns

• Supported: Stale 'dist/' resolution: If 'pi-mono' source entrypoints are launched from outside the directory, 'tsx' may resolve stale 'dist/' packages instead of source, causing new models to appear missing. Fix: Force 'tsconfig' in test scripts. 2. Keychain Auth: Credential resolution may fail if the wrong keychain file or search scope is used. 3. Bedrock 4.7: Availability and adaptive-thinking semantics on Amazon Bedrock are unverified and deferred. 4. Gemma Structured Output: Gemma models may wrap JSON in markdown fences, requiring harness adjustments for strict schema adherence. Evidence: README.md; artifacts/handoffs/2026-04-17-claude-opus-4-7-execution-plan.md; artifacts/handoffs/2026-04-03-gemma4-structured-output-probe-handoff.md.

Next Best Action

• Partially supported: Execute Track B of the Opus 4.7 plan: Implement Keychain-first credential resolution. 2. Verify Bedrock 4.7 availability and semantics if needed for future default swaps. 3. Continue monitoring the builder swarm dispatch loop for stability. 4. Consider running the Gemma 4 benchmark suite if hardware resources allow, to evaluate potential replacement of Qwen 3.5. Evidence: artifacts/handoffs/2026-04-17-claude-opus-4-7-execution-plan.md; artifacts/handoffs/2026-04-03-gemma4-node-update-handoff.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

14. rolo-local

written 2026-05-19T00:30:31Z · verified stale (detector 2026-06-15T21:41:04Z)

• project_id: rolo-local
• priority_rank: 14
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:30:31Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project provides a local-inference variant of the Rolo co-founder system, enabling Dr. Andrew to run Rolo on local hardware (NODE1, NODE2, Mac mini) via Ollama. It features a slot-based architecture for routing inference to specific models (e.g., qwen3.6, nemotron3, gemma4) and nodes. Recent work has focused on stabilizing the local backend, integrating pcsl-ontology v0, and refining the boot preflight protocol to ensure memory continuity. Evidence: README.md; AGENTS.md; artifacts/handoffs/2026-04-22-rolo-cloud-refresh-after-ontology-v0.md.

Current State

• Supported: As of 2026-04-28, the system is operational with a slot-based architecture. Slot 6 (default) uses qwen3.6:27b on NODE1. Slot 1 uses nemotron3:33b on NODE1. Slot 11 uses gemma4:e4b on the Mac mini. The pcsl-ontology v0 is installed and exposed via CLI. The boot preflight protocol is self-executing via rolo-boot.sh. Shared memory (pcsl-context) is synchronized with cloud Rolo. Evidence: README.md; AGENTS.md; artifacts/handoffs/2026-04-22-rolo-cloud-refresh-after-ontology-v0.md.

Risks, Blockers, And Unknowns

• Partially supported: SSH tunnel dependency: Local inference requires active SSH tunnels to NODE1/NODE2. If tunnels drop, local slots fail. 2. Model availability: Ollama tags must be present on nodes; missing models cause slot failures. 3. Memory degradation: If rolo-boot.sh fails or pcsl-context is unavailable, memory state is degraded. 4. Git usage: The project is not a git repo; version control is not active, which may impact change tracking. Evidence: README.md; AGENTS.md; artifacts/handoffs/2026-03-29-llm-ide-preflight-fix.md.

Next Best Action

• Partially supported: Verify SSH tunnels to NODE1 and NODE2 are active. 2. Run rolo-boot.sh to confirm memory and graph state are healthy. 3. Check slot 6 (default) and slot 1 (nemotron3) availability via rolo-cli.sh --local-probe. 4. Review pcsl-ontology v0 health via rolo-cli.sh --ontology health. 5. Ensure no drift between local and cloud Rolo memory states. Evidence: README.md; AGENTS.md; artifacts/handoffs/2026-04-22-rolo-cloud-refresh-after-ontology-v0.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

15. dewdrop

written 2026-05-19T00:31:30Z · verified fresh (detector 2026-06-27T00:58:21Z)

• project_id: dewdrop
• priority_rank: 15
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:31:30Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project is currently at 'hatched v1' stage with no implemented runtime. It functions as a compact assistant identity complementing 'Rolo' with quiet precision and scope discipline. The repository contains core records, avatar assets pending approval, and review contracts for evidence and privacy. Evidence: README.md; AGENTS.md; pcsl-gen-status-dewdrop.md.

Current State

• Supported: Stage: hatched v1. Runtime: not implemented yet. Avatar: v1 candidate saved in assets, final likeness approval pending. Codex app pet installed at [redacted-local-path] Priority rank 15 with roster status 'active'. Evidence: README.md; pcsl-gen-status-dewdrop.md.

Risks, Blockers, And Unknowns

• Supported: Medium severity risk/blocker identified regarding avatar v1 assets, prompt, checksums, and acceptance boundary in assets/avatar/README.md. Runtime health, readiness, and deployment state remain unknown as runtime is not implemented. Evidence: README.md; pcsl-gen-status-dewdrop.md.

Next Best Action

• Supported: Review the cited evidence (README.md, AGENTS.md), then refresh or add a closeout/handoff artifact if the active objective is stale or incomplete. Confirm avatar likeness approval with Andrew. Evidence: README.md; AGENTS.md; pcsl-gen-status-dewdrop.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

16. pcsl-rolo-builder

No evidence · freshness: unavailable — no detector signal · written 2026-05-19T00:31:54Z

• project_id: pcsl-rolo-builder
• priority_rank: 16
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:31:54Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: No evidence
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project provides a suite of shell scripts to manage a G0-G11 readiness pipeline, run local dyads (Slot 6 alpha -> omega), validate outputs, and report status. It relies on local artifacts for proof and audit, specifically avoiding external network probes unless explicitly approved. The current evidence base is sparse, consisting primarily of the README and status metadata. Evidence: README.md; pcsl-gen-status-pcsl-rolo-builder.md.

Current State

• Partially supported: The project is marked as 'active' in the priority list. The README indicates that G0.5.strategy is PASS, G4.power_readiness is PASS for the latest Phase 5f run, and G9.5.llm_res_smoke is SKIPPED by policy. However, runtime health and deployment state are unknown as no runtime probes were performed in this evidence packet. Evidence: README.md; data/project-priority-list.json; pcsl-gen-status-pcsl-rolo-builder.md.

Risks, Blockers, And Unknowns

• Unknown: Runtime health, deployment state, and current human priorities outside the reviewed files are unverified. No specific blockers were identified in the approved evidence, but the sparse evidence base limits confidence in the operational state. Evidence: no direct packet citation; classification remains Unknown.

Next Best Action

• Partially supported: Add or approve a safe README, closeout, or handoff file that fits project_biopsy_sources, then rerun the v0.0.4 status writer to improve evidence density. Evidence: pcsl-gen-status-pcsl-rolo-builder.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

17. grok-sage

written 2026-05-19T00:32:18Z · verified fresh (detector 2026-06-27T01:00:50Z)

• project_id: grok-sage
• priority_rank: 17
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:32:18Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project consists of two distinct layers: 'grok-sage.py' (the primary assistant runtime using xAI) and 'pi_grok' (a deterministic repo-awareness layer). 'pi_grok' scans repositories into static context files or delegates live queries to a separate subagent (currently configured for Anthropic models via pi-mono). Recent plans focus on fixing outer timeout bugs related to subagent fallbacks, swapping the primary subagent model from Opus to Sonnet for speed/cost efficiency, adding self-awareness for the grok-sage repo itself, and expanding input capabilities with prompt_toolkit. Evidence: README.md; docs/plans/outer-timeout-budget-fix.md; docs/plans/subagent-model-swap.md; docs/plans/grok-sage-self-awareness.md; docs/plans/prompt-toolkit-input.md.

Current State

• Supported: The project is active (priority rank 17). The core CLI is functional with support for interactive sessions, one-shot queries, and built-in xAI tools. The 'pi_grok' subagent system is implemented with Phase 1 (static scanning) and Phase 2 (live RPC) capabilities. Currently, 'pcsl-builder' is the primary indexed repo for subagent grounding. The subagent configuration still lists 'claude-opus-4-6' as primary in documentation/examples, though plans indicate a swap to 'claude-sonnet-4-6' is pending or in progress to address timeout issues. Evidence: pcsl-gen-status-grok-sage.md; README.md; docs/plans/subagent-model-swap.md; docs/plans/outer-timeout-budget-fix.md.

Risks, Blockers, And Unknowns

• Supported: Outer Timeout Race Condition: The current outer timeout (150s) kills the subprocess while the fallback model is still working, leading to opaque failures. 2. Subagent Latency: Opus is too slow for subagent queries, causing frequent timeouts. 3. Scanner Blind Spots: The scanner does not currently read 'pyproject.toml' or '.grok-sage/*.md' identity files, resulting in thin context for Python repos like grok-sage itself. 4. Self-Scan Circularity: Scanning grok-sage requires updating SKIP_DIRS to prevent circular inclusion of '.pi-grok/context/'. Evidence: docs/plans/outer-timeout-budget-fix.md; docs/plans/subagent-model-swap.md; docs/plans/grok-sage-self-awareness.md.

Next Best Action

• Supported: Implement the outer timeout budget fix in 'grok-sage.py' to prevent premature subprocess termination during fallback. Simultaneously, update '.env' to swap the primary subagent model to 'claude-sonnet-4-6' and fallback to 'claude-haiku-4-5-20251001' as per the model swap plan. Verify changes with 'pytest' and smoke tests. Evidence: docs/plans/outer-timeout-budget-fix.md; docs/plans/subagent-model-swap.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

18. pcsl-ui

written 2026-05-19T00:33:15Z · verified fresh (detector 2026-06-27T01:09:28Z)

• project_id: pcsl-ui
• priority_rank: 18
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:33:15Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project is a React-based frontend application (v0.1.0) built with Vite and TypeScript. It implements a specific visual theme using 'Press Start 2P' fonts and PCSL brand colors (Lime, Cyan, Magenta). The application includes custom components for navigation (NintendoCursor), transitions (QuantumTransition), and contact (EmailButton). It is currently in an active state with a local development server running on port 3020. Evidence: README.md; package.json; pcsl-gen-status-pcsl-ui.md.

Current State

• Supported: The project is active (rank 18) and version 0.1.0. It is a private npm package (@pcsl/ui) using ES modules. The codebase includes React 18, React Router, Tailwind CSS, and TypeScript. Local development is configured to run on port 3020. Production builds are optimized for AWS Amplify deployment. Evidence: package.json; README.md; pcsl-gen-status-pcsl-ui.md.

Risks, Blockers, And Unknowns

• Unknown: Runtime health, readiness, and deployment state are unverified. The previous status generation noted insufficient evidence for project role inference and that runtime behavior was not tested. No specific blockers were identified in the approved evidence. Evidence: pcsl-gen-status-pcsl-ui.md.

Next Best Action

• Partially supported: Verify local development environment by running 'npm run dev' and accessing [redacted-url] Consider implementing the listed future enhancements (sound effects, mobile touch support) if prioritized. Ensure production builds are tested for AWS Amplify deployment. Evidence: README.md; pcsl-gen-status-pcsl-ui.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

19. pcsl-console

written 2026-05-19T00:33:40Z · verified fresh (detector 2026-06-27T01:02:13Z)

• project_id: pcsl-console
• priority_rank: 19
• roster_status: active
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:33:40Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project has evolved from a simple CLI chat with notes (v0-v1) to a complex, A2A-compliant multi-agent runtime with a persistent memory graph, skill packs, voice adapters, and a React/Electron desktop console (v6). It supports streaming responses, semantic search, and structured UI artifacts. The codebase is Python-based (>=3.11) with optional A2A dependencies. Evidence: README.md; pyproject.toml.

Current State

• Partially supported: The project is at version 4.0.0 in pyproject.toml, but the README documents features up to v6 (Console Foundation, Electron App, A2UI Compatibility). The v5B build records indicate v5A (Memory Graph) was complete with 275 passing tests. The current state appears to be a mature, feature-rich local agent runtime with a desktop UI, though the exact integration status of v6 features is not fully detailed in the provided excerpts. Evidence: pyproject.toml; README.md; docs/records/PCSL-MCP-docs/records-PCSL-MCP/cursor_v5b.md.

Risks, Blockers, And Unknowns

• Unknown: The exact status of v6 feature integration (Console Foundation, A2UI Compatibility) is not fully clear from the excerpts. Potential risks include complexity in maintaining multiple agent versions, A2A protocol compliance across different clients, and performance of semantic search with large memory graphs. Evidence: no direct packet citation; classification remains Unknown.

Next Best Action

• Partially supported: Verify the current state of the v6 features in the desktop console. Ensure all skill packs are properly loaded and tested. Review the A2A server configuration for any security concerns (e.g., host binding). Consider documenting any remaining gaps in the v6 feature set. Evidence: README.md; docs/records/PCSL-MCP-docs/records-PCSL-MCP/cursor_v5b.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

20. plsrev0

written 2026-05-19T00:34:34Z · verified fresh (detector 2026-06-27T01:19:24Z)

• project_id: plsrev0
• priority_rank: 20
• roster_status: sandbox
• status_file_present: true
• status_file_schema: pcsl-gen-status-v1-local-private
• classification: local_private
• redaction_required_for_consolidation: true
• machine_authority: false
• model: qwen3.6:27b
• last_update_utc: 2026-05-19T00:34:34Z
• freshness: current
• validation: pass
• redaction: clean
• evidence_label: Supported
• ingestion_status: approved_and_present

Review Summary

Executive Summary

• Supported: The project is approximately 70% complete. Core infrastructure including the KERNAL orchestration engine, INTERCHANGE API gateway, and five AI provider integrations are established. Foundation processing (ktech1) is complete, while RAG (ktech2) is partially implemented via AWS Bedrock. Recent work has focused on dynamic service discovery to eliminate hardcoded ports, implementing a robust RBAC system for an admin console, and fixing real-time data capture in the TUNE optimization service. The platform supports HIPAA-compliant workflows and includes a clinical triage system. Evidence: README.md; docs/DYNAMIC_SERVICE_DISCOVERY_GUIDE.md; docs/RUN2_DAY1_SUMMARY.md; docs/TUNE_REALTIME_FIX.md.

Current State

• Partially supported: The system is in active development with a functional local development environment. Key services include the Frontend (port 5173), INTERCHANGE API (port 3000), Foundation Endpoints (ports 3003-3007), PCSL Virtual (port 5175), TUNE (port 3008), and Health Console (port 3002). The Admin Console (port 3300) and PCSL EMU (port 3021) are integrated. Dynamic service discovery via Consul is being implemented to replace hardcoded ports. RBAC infrastructure is 87% complete, and the Admin UI is 100% complete. Evidence: README.md; docs/RUN2_ADMIN_INTEGRATION_SUMMARY.md; docs/RUN2_DAY1_SUMMARY.md; docs/RUN2_DAY2_SUMMARY.md.

Risks, Blockers, And Unknowns

• Unknown: Specific production deployment timelines and current funding status are not detailed in the provided evidence. Potential blockers include the completion of MCP/A2A/RLHF implementations and the finalization of dynamic port migration for all services. The health endpoint for the Admin Console currently returns 404, indicating a middleware issue that needs debugging. Evidence: docs/RUN2_ADMIN_INTEGRATION_SUMMARY.md.

Next Best Action

• Partially supported: Debug and fix the Admin Console health endpoint middleware to resolve the 404 error. 2. Continue migrating remaining services to dynamic port discovery using Consul. 3. Test the TUNE real-time data capture by running a full triage session and exporting the generated training data. 4. Complete the integration of the 'useRole' hook with the actual authentication context in the Admin UI. Evidence: docs/RUN2_ADMIN_INTEGRATION_SUMMARY.md; docs/DYNAMIC_SERVICE_DISCOVERY_GUIDE.md; docs/TUNE_REALTIME_FIX.md; docs/RUN2_DAY1_SUMMARY.md.

Manual Review Prompts

• Is the project purpose accurate against the cited source files?
• Is the current state stale, overstated, or missing a recent closeout?
• Are the risks, unknowns, and next safe action useful enough for operator handoff?
• Should any local-private detail be tightened before broader display?

21. pcsl-hermes

written UNKNOWN — status file pending · verified fresh (detector 2026-06-27T01:34:12Z)

• project_id: pcsl-hermes
• priority_rank: 21
• roster_status: active
• status_file_present: false
• status_file_schema: null
• classification: unspecified
• redaction_required_for_consolidation: null
• machine_authority: null
• model: null
• last_update_utc: null
• freshness: missing
• validation: missing
• redaction: not_run
• evidence_label: Unknown
• ingestion_status: not_yet_approved

22. pcsl-hub

written UNKNOWN — status file pending · verified fresh (detector 2026-06-27T01:06:16Z)

• project_id: pcsl-hub
• priority_rank: 22
• roster_status: active
• status_file_present: false
• status_file_schema: null
• classification: unspecified
• redaction_required_for_consolidation: null
• machine_authority: null
• model: null
• last_update_utc: null
• freshness: missing
• validation: missing
• redaction: not_run
• evidence_label: Unknown
• ingestion_status: not_yet_approved

23. pcsl-identify

written UNKNOWN — status file pending · verified stale (detector 2026-06-15T20:32:40Z)

• project_id: pcsl-identify
• priority_rank: 23
• roster_status: active
• status_file_present: false
• status_file_schema: null
• classification: unspecified
• redaction_required_for_consolidation: null
• machine_authority: null
• model: null
• last_update_utc: null
• freshness: missing
• validation: missing
• redaction: not_run
• evidence_label: Unknown
• ingestion_status: not_yet_approved

24. pcsl-pi

written UNKNOWN — status file pending · verified fresh (detector 2026-06-27T00:54:06Z)

• project_id: pcsl-pi
• priority_rank: 24
• roster_status: active
• status_file_present: false
• status_file_schema: null
• classification: unspecified
• redaction_required_for_consolidation: null
• machine_authority: null
• model: null
• last_update_utc: null
• freshness: missing
• validation: missing
• redaction: not_run
• evidence_label: Unknown
• ingestion_status: not_yet_approved