Capablio

Handbook

Feature status

Single source of truth for what the feedback-360 system does today. Every capability carries exactly one flag:

Updated

  • Implemented — shipped, verified by the orchestration checks and test suite.
  • Partial / Unimplemented — used only on the Adaptive Feedback Framework track (AFF-*); see framework/implementation-status.md.

The multi-user forge360 survey platform (all PLN-* rows below) is implemented alongside the offline pipeline and Studio. Engineering docs: forge360/README.md. Original product spec: survey-platform-vision.md.

Production-scale next track: Capablio Adaptive Feedback Framework v0.1 — architecture under framework/README.md; all new product capabilities are tracked as AFF-* with explicit Unimplemented / Partial flags until shipped.

The reader-facing mirror of this page is report/wiki/roadmap.html, regenerated by the render stage (dual-wiki convention — see app-logic.md).

Implemented today

ID Capability Status Doc
IMP-01 Offline pipeline: Thomas 360 PDF (+ optional snapshot deck) → data.json → metrics.json → interactive HTML report Implemented pipeline.md
IMP-02 Deterministic analytics: quadrants, self gaps, priority scores, sentiment, SWOT, section summaries (no ML, no network) Implemented pipeline.md
IMP-03 Verification harness: arithmetic reconciliation, independent PDF re-read, golden drift gate Implemented pipeline.md
IMP-04 360 Insights Studio: operator orchestration (KB, jobs board, Capablio gateway, ingestion) Implemented app-logic.md
IMP-05 Verify-gated export of the offline zip pack Implemented app-api.md
IMP-06 Local Studio HTTP API on 127.0.0.1 (no auth, single-user) Implemented app-api.md
IMP-07 PDP guided interview with evidence pushback, override reasons, finalized pdp.json + trackable checklist Implemented pdp.md
IMP-08 Optional AI narrative enrichment with in-memory-only provider keys Implemented ai-providers.md
IMP-09 Live weight controls recomputing metrics via /api/metrics Implemented README.md
IMP-10 Reader wiki generated with every render (this page's HTML mirror included) Implemented pipeline.md
IMP-11 Desktop Electron shell and packaged installers Implemented desktop.md, packaging.md

forge360 survey platform (implemented)

All rows below are shipped in the forge360/ FastAPI package. The offline pipeline and Studio remain the local single-user tools; forge360 adds authenticated multi-user survey collection, encryption, wrap-up PDFs, and subject PDP.

1. Identity & access

ID Capability Status
PLN-ID-01 Sign-in with Google; role-based authorization Implemented
PLN-ID-02 Dual hierarchy model: HR hierarchy + Org hierarchy drive permissions (single demo tenant; multi-tenant scoping via PLN-TEN-*) Implemented
PLN-ID-03 Only the subject and their direct HR can access the subject's personal data Implemented
PLN-ID-04 HR levels above the direct HR see only aggregate analytics for their designated organizations (same scope as managers) Implemented
PLN-ID-05 Managers see only GDPR/EU-regulation-compatible information (aggregates, completion, no personal detail) Implemented

2. REST API

ID Capability Status
PLN-API-01 API-first REST design with bearer-token authorization on every endpoint Implemented

3. Survey lifecycle

ID Capability Status
PLN-SRV-01 HR initiates period surveys (linked to a quarter) for their org structures Implemented
PLN-SRV-02 HR and managers see survey completion status for their teams Implemented
PLN-SRV-03 System generates per-person invite email text (outlining the process) for HR to send manually Implemented
PLN-SRV-04 System generates the results-access email for the subject after HR confirms the survey was communicated; HR adds attachments manually Implemented

4. Conversational survey UI

ID Capability Status
PLN-UI-01 Chat-style respondent experience with progress indicator, facilitating conversation while collecting responses Implemented
PLN-UI-02 Collects the same data as today's PDFs (feedback + snapshot) plus richer free-form information Implemented

5. Privacy & encryption

ID Capability Status
PLN-ENC-01 Rich respondent data encrypted so that only the direct HR and the subject can decrypt; system operators cannot Implemented
PLN-ENC-02 PDP encrypted so only the subject can read the actual plan data Implemented
PLN-ENC-03 Zero-knowledge posture: a person with physical access to the database and the system cannot extract personal data Implemented
PLN-ENC-04 Generated PDFs are never stored by the system Implemented

6. Wrap-up & PDP lifecycle

ID Capability Status
PLN-PDP-01 On survey completion, results are wrapped into a feedback PDF + snapshot PDF, available only to the direct HR as a one-time download Implemented
PLN-PDP-02 Subject gets system access to work with survey results, generate a PDP, and store it (encrypted) in the system Implemented
PLN-PDP-03 Subject works with the PDP throughout the period and marks completed goals Implemented
PLN-PDP-04 HR and managers see only general statistics about activity inside the system Implemented

forge360 multi-tenant SaaS (PLN-TEN-*)

ID Capability Status
PLN-TEN-01 Multiple customer tenants on one deployment with data isolation Implemented
PLN-TEN-02 Platform operator creates and suspends tenants Implemented
PLN-TEN-03 Tenant admin manages org unit tree within tenant Implemented
PLN-TEN-04 Tenant admin assigns HR designations (HR user ↔ org unit) Implemented
PLN-TEN-05 Tenant admin creates/edits users (org unit, manager, HR parent, roles) Implemented
PLN-TEN-06 Invite flow: pre-provision by email; link on first Google sign-in Implemented
PLN-TEN-07 CSV bulk user import (tenant-wide and per org-unit node) with validation preview Implemented
PLN-TEN-08 All existing survey/crypto APIs tenant-scoped (no cross-tenant leakage) Implemented

forge360 entry path — PDF import (PLN-ENTRY-*)

ID Capability Status
PLN-ENTRY-01 Subject or direct HR uploads external Thomas/snapshot PDFs and enters PDP journey Implemented
PLN-ENTRY-02 Import creates pdf_import survey in wrapped with sealed results Implemented

forge360 KB enrichment (PLN-KB-*)

ID Capability Status
PLN-KB-01 Operator API to run/poll/cancel category enrichment Implemented
PLN-KB-02 LCDL .lmeta flow (kb-category-enrich) via ChatGPT CDP Implemented
PLN-KB-03 Merge/preview promoted fragments into KB pack Implemented
PLN-KB-04 Studio KB admin hierarchy browse (tree API + browse_tree index) Implemented
PLN-KB-05 Deterministic fragment validation gate Implemented

forge360 Capablio UI extensions (PLN-UI-*)

ID Capability Status
PLN-UI-03 Google Sign-In in /app (tenant-aware) Implemented
PLN-UI-04 Client key registration and PDF import UI Implemented
PLN-UI-05 Subject results view and PDP chat in /app Implemented
PLN-UI-06 Combined organization admin (/app/admin/: directory, CSV, invites) Implemented
PLN-UI-07 HR completion overview and survey lifecycle UI in /app/admin/ Implemented
PLN-UI-08 Platform operator tenant console (/app/operator/) Implemented
PLN-UI-09 End-user account & security (key backup, reseal) in /app/ Implemented
PLN-UI-10 Invite-flow encryption key import on sign-in Implemented
PLN-UI-11 Kitchen Sink enterprise console shell (forge-capablio-console.css, Capablio brand) Implemented
PLN-UI-12 KS product_page app shell (forge-capablio-app-shell.css, forge-sidebar, data-fs-pack="enterprise") on /app, /app/operator/, /app/admin/ Implemented

forge360 crypto extensions (PLN-ENC-*)

ID Capability Status
PLN-ENC-05 Automated zero-knowledge audit in deploy gate Implemented
PLN-ENC-06 Envelope-only decrypt endpoints (no private_key_b64 to server) Implemented

forge360 worker gateway (PLN-WKR-*)

ID Capability Status
PLN-WKR-01 Remote job broker (RemoteJob) with operator and worker APIs Implemented
PLN-WKR-02 Local capablio-worker gateway (KB enrich, pipeline, AI, Cursor) Implemented
PLN-WKR-03 Cloudflare Tunnel exposure + deploy secrets for worker URL/token Implemented
PLN-WKR-04 Studio orchestrator UI (/orchestrator) for jobs and gateway health Implemented
PLN-WKR-05 Capablio user parity: results HTML, verify, export zip, wiki, live weights Implemented

Adaptive Feedback Framework (AFF-*) — production scale

Do not mark these as Implemented until code, tests, and docs land together. Full matrix: framework/implementation-status.md.

AFF flag totals (2026-07-17): 46 Implemented · 0 Partial · 0 Unimplemented (46 AFF-* rows).

Area IDs Status summary
Question architecture (family/variant, packs) AFF-QA-* (6) Implemented
Cadence & exposure cooldown (shared ledger) AFF-CD-* (8) Implemented
Survey assembly & coverage engine AFF-AS-* (7) Implemented
PDP adaptive recurrence AFF-PDP-* (5) Implemented (baseline PDP is PLN-PDP-*)
Tenant overlays & pack governance AFF-TN-* (4) Implemented
Thomas import levels & evidence bridge AFF-TH-* (5) Implemented
Analytics, privacy ops, integrations AFF-OP-* (8) Implemented
Subject-seat entitlements AFF-EN-* (3) Implemented

Flag governance

  • New platform capabilities (PLN-*) start as planned rows and move to Implemented when code, tests, and docs land together.
  • New AFF capabilities start as Unimplemented (or Partial) in framework/implementation-status.md and move to Implemented only when shipped.
  • When any row changes, update both this file (or the AFF matrix) and _wiki_roadmap in feedback360/render_html.py, then regenerate the render stage.