mptyl/ThothII
1
0
Fork 0
Code Issues 25 Pull Requests Actions Packages Projects Releases 1 Wiki Activity

docs: record P7 publication + live stack (pending VPN)

Browse Source
This commit is contained in:
marcopan
2026-08-13 16:41:46 +02:00
parent 3046ac34c6
commit 84b233d937
3 changed files with 75 additions and 102 deletions
Download Patch File Download Diff File Expand all files Collapse all files
PROJECT_STATE.md
+15 -7
View File
@@ -7,7 +7,7 @@
> ThothII per il repository (app + CLI `thothctl`), (3) come usare l'applicazione ThothII di base
> (sessioni, domande, gate). Il documento userà parole semplici ed esempi; i dettagli tecnici
> resteranno nei contratti esistenti. Esempio pratico completo: Policlinico San Donato.
Last updated: 2026-08-13 (P2–P6 accepted; aggregate PASS; P7 repo restructured + validated).
Last updated: 2026-08-13 (P2–P6 accepted; P7 published + live stack, pending VPN).
> Point a fresh session here ("read PROJECT_STATE.md") before substantial work.
### P3 effective configuration and `.tht-dwh` — implementation complete, automated PASS, manual PASS (2026-08-13)
@@ -146,12 +146,20 @@
- **Templates:** `deploy/psd/{workspace-bindings,operator,thothii-installation}.env.example` +
gitignored `secrets/`; operator checklist in `docs/install/psd-workspace-setup.md` (registered in
MkDocs nav).
- **Owner-gated (needed to continue):** (1) GitHub remote URL + push credentials for the PSD repo;
(2) DWH X-API-Key (+ CA) reuse; (3) LLM provider/model confirmation (default `zai/glm-5.2` already
in the descriptor); (4) confirmation of the DWH REST ping (`POST /rpc/ping`, `X-API-Key`) for
`diagnostics.dwh_rest`.
- **Remaining:** Tasks 3–5 of the plan (installation bindings/secrets, stack start +
`qwen3-embedding:0.6b` pull, re-embedding/`preprocess run`, live P8 L2 smoke).
- **Published (2026-08-13):** private repo `https://github.com/mptyl/tht-workspace-psd` (main =
`d4f9185`), consumed via SSH deploy key `thothii-psd` (read-write, passphrase-less, generated in
`deploy/psd/secrets/`). Real operator config is wired (gitignored): `deploy/psd/operator.env`,
`workspace-bindings.env`, `thothii-installation.yaml`, `connector-secrets.yaml` + `secrets/`
(DWH X-API-Key reused from the legacy `.env`; no CA — the DWH REST is public HTTPS).
- **Stack live:** started via `thothctl start` (project `thothii-70417a3e30ea`), all services
healthy, `qwen3-embedding:0.6b` present; the registry cloned + activated `psd-clinical`
(`ready`); `thothctl workspace inspect` returns `ok` with descriptor/catalog/runtime identities.
Gotcha recorded: `thothctl` uses a per-descriptor Compose project name, so the stack must be
started with `thothctl start` (not a raw `compose-with-preflight.sh up`).
- **Blocker:** VPN/DNS — `supabase-aritmolab.policlinicosandonato.it` does not resolve (NXDOMAIN),
so DWH `preprocess run` and the live P8 L2 session are pending VPN.
- **Remaining:** activate VPN → `thothctl … workspace preprocess run --workspace psd-clinical` →
live session smoke (P8 L2).
### Final aggregate P2–P6 verification — automated PASS, manual PENDING (2026-08-13)
deploy/psd/operator.env.example
+21 -7
View File
@@ -1,16 +1,30 @@
# Copia in deploy/psd/operator.env (non tracciato). Solo path non-segreti.
THT_WORKSPACE_GIT_REMOTE=<URL-del-repo-github-del-workspace>
THT_WORKSPACE_GIT_REMOTE=git@github.com:mptyl/tht-workspace-psd.git
THT_WORKSPACE_GIT_BRANCH=main
THT_WORKSPACE_INSTALLATION_ID=psd-local
THT_WORKSPACE_GIT_AUTHOR_NAME="Thoth PSD"
THT_WORKSPACE_GIT_AUTHOR_EMAIL=thoth-psd@example.invalid
THT_WORKSPACE_BINDINGS_ENV_FILE=<abs>/deploy/psd/workspace-bindings.env
THT_WORKSPACE_GIT_SSH_KEY_FILE=<abs>/deploy/psd/secrets/git-ssh-key
THT_WORKSPACE_GIT_KNOWN_HOSTS_FILE=<abs>/deploy/psd/secrets/git-known-hosts
# Credenziali/secrets: path HOST (mai il contenuto).
# App
THT_SECRETS_FILE=<abs>/deploy/psd/secrets/thothii.secrets
PI_AUTH_FILE=<abs>/deploy/psd/secrets/pi-auth.json
THT_WORKSPACE_BINDINGS_ENV_FILE=<abs>/deploy/psd/workspace-bindings.env
# Connector secret sources (host-only paths)
THT_WS_PSD_CLINICAL_DWH_API_KEY_SOURCE=<abs>/deploy/psd/secrets/psd-clinical-dwh-api-key
# Opzionale (solo con CA interna):
THT_WS_PSD_CLINICAL_DWH_TLS_CA_SOURCE=<abs>/deploy/psd/secrets/psd-clinical-dwh-ca.pem
# Git HTTPS privato:
THT_WORKSPACE_GIT_CREDENTIALS_FILE=<abs>/deploy/psd/secrets/git-credentials
# Pi (LLM)
PI_PROVIDER=zai
PI_MODEL=glm-5.2
PI_THINKING=medium
# App defaults
AUTH_MODE=none
THOTH_PUBLIC_EXPOSURE=false
MAX_PI_PROCESSES=4
THOTH_HTTP_PORT=8080
THOTH_CORE_HTTP_PORT=8787
THT_DB_NAME=postgres
THT_DWH_REST_URL=https://supabase-aritmolab.policlinicosandonato.it/dwh/
docs/install/psd-workspace-setup.md
+39 -88
View File
@@ -1,107 +1,58 @@
# Policlinico San Donato — setup workspace (nuova gestione)
Guida operativa per collegare ThothII al DWH di PSD con il nuovo sistema (registry Git + descriptor
v3 + `thothctl`). Lo stato attuale:
v3 + `thothctl`).
- il repository PSD **è già stato ristrutturato** e committato localmente in
`/Users/mp/projects/tht-workspace-psd` (layout P1.1: `thoth-workspaces.yaml` +
`psd-clinical/workspace.yaml` + `psd-clinical/evidence/` + `psd-clinical/schema/annotations.yaml`);
- la **validazione locale** del registry è già passata (descriptor v3 valido, 36 Evidence
materializzate, 42 KB di annotations sincronizzate, docs generate).
## Stato attuale (2026-08-13)
Rimangono i passi che richiedono segreti/accesso. I template pronti sono in `deploy/psd/`
(`*.example`).
- **Repository PSD pubblicato:** `https://github.com/mptyl/tht-workspace-psd` (privato), branch
`main`, commit `d4f9185`. Layout P1.1 già migrato e validato.
- **Deploy key SSH** (read-write, senza passphrase) generata in
`deploy/psd/secrets/git-ssh-key` e registrata sul repo come deploy key `thothii-psd`; il remote
Git usato dall'installazione è `git@github.com:mptyl/tht-workspace-psd.git`.
- **Config operatore pronta** (file reali gitignored in `deploy/psd/`): `operator.env`,
`workspace-bindings.env`, `thothii-installation.yaml`, `connector-secrets.yaml` e i secret
`secrets/` (API key DWH riusata, pi-auth, secret bundle, chiave SSH, known_hosts). Nessuna CA:
il DWH REST usa HTTPS pubblico (`THT_SSL_CA` era vuoto).
- **Stack avviato** (progetto `thothii-70417a3e30ea`, via `thothctl start`): `qdrant`, `embedding`
(con `qwen3-embedding:0.6b`), `core`, `frontend` sani. Il registry ha **clonato e attivato**
`psd-clinical` (stato `ready`).
- **`thothctl workspace inspect --workspace psd-clinical` = OK** (identità descrittore/catalogo/
config risolte con i bindings DWH).
- **Bloccante residuo: VPN.** `supabase-aritmolab.policlinicosandonato.it` non risolve
(`NXDOMAIN`) → il preprocessing DWH e le sessioni live non possono ancora partire.
## 1. Pubblica il repository PSD su GitHub
## Avvio/arresto (canonico)
Crea un repository Git **privato** su GitHub e pubblicalo:
Usare `thothctl` (stesso project name, quindi stessi volumi named):
```bash
git -C /Users/mp/projects/tht-workspace-psd remote add origin <URL-repo>
git -C /Users/mp/projects/tht-workspace-psd push -u origin main
THOTHCTL=dist/thothctl/thothctl-darwin-arm64
"$THOTHCTL" --installation "$(pwd)/deploy/psd/thothii-installation.yaml" start
"$THOTHCTL" --installation "$(pwd)/deploy/psd/thothii-installation.yaml" workspace inspect --workspace psd-clinical --json
"$THOTHCTL" --installation "$(pwd)/deploy/psd/thothii-installation.yaml" stop
```
Oppure fornisci URL + credenziali e lo pusho io.
> **Nota project name:** `thothctl` calcola un project name stabile dall'installation descriptor
> (`thothii-<hash>`); `docker compose` "a mano" usa invece `name: thothii` dal `compose.yaml`, quindi
> i volumi named non coinciderebbero. Perciò per lo stack si usa `thothctl start` (non
> `compose-with-preflight.sh up`).
## 2. Prepara i segreti (in `deploy/psd/secrets/`, non tracciati)
| File | Contenuto | Riuso |
| --- | --- | --- |
| `psd-clinical-dwh-api-key` | la `X-API-Key` del DWH REST | `THT_DWH_API_KEY` esistente |
| `psd-clinical-dwh-ca.pem` | CA interna (solo se il REST la presenta) | `THT_SSL_CA` esistente |
| `pi-auth.json` | profilo Pi con le credenziali LLM (zai/deepseek/qwen) | esistente |
| `thothii.secrets` | bundle secret dell'app | esistente |
| `git-credentials` | credenziali HTTPS del remote Git privato | nuova |
> **Nota CA**: il *preprocessing* DWH via REST supporta la CA (`TLS_CA_FILE`). Il *diagnostic* di
> readiness REST, invece, rifiuta una CA per-request privata: se il ping fallisce per la CA, va
> reso trusted a runtime oppure si adatta `diagnostics.dwh_rest` nel descriptor Git.
## 3. Compila i file operatore (dai template)
## Completare quando la VPN è attiva
```bash
cd deploy/psd
cp workspace-bindings.env.example workspace-bindings.env
cp operator.env.example operator.env
cp thothii-installation.yaml.example thothii-installation.yaml
"$THOTHCTL" --installation "$(pwd)/deploy/psd/thothii-installation.yaml" workspace preprocess run --workspace psd-clinical --json
```
Compila i placeholder con i valori reali (remote GitHub, path assoluti). Il bindings file per PSD è:
Se si ferma per le join (`manual_review_required`): rivedere `psd-clinical/schema/annotations.yaml`
nel repo, commit+push, poi `workspace schema accept --run <run-id> --yes` e
`workspace preprocess run --resume <run-id>`.
```dotenv
THT_WS_PSD_CLINICAL_DWH_TRANSPORT=rest_api
THT_WS_PSD_CLINICAL_DWH_BASE_URL=https://supabase-aritmolab.policlinicosandonato.it/dwh/
THT_WS_PSD_CLINICAL_DWH_API_KEY_FILE=/run/secrets/psd-clinical-dwh-api-key
```
Poi aprire `http://localhost:8080`, selezionare `psd-clinical` e creare una sessione con una domanda.
## 4. Genera il connector override
## Cosa è già stato fatto
```bash
scripts/generate-connector-secrets-override.sh --bindings-env deploy/psd/workspace-bindings.env --operator-env deploy/psd/operator.env --output deploy/psd/connector-secrets.yaml --service workspace-maintenance --role all
```
## 5. Avvia lo stack (VPN attiva; internet per scaricare il modello)
```bash
scripts/compose-with-preflight.sh --env-file deploy/psd/operator.env -f compose.yaml -f deploy/compose.local.yaml -f deploy/compose.git-https.yaml -f deploy/psd/connector-secrets.yaml up --build -d
```
Il primo avvio esegue `embedding-model-init`, che scarica `qwen3-embedding:0.6b` nel volume
`embedding-models` e poi sblocca `core`.
## 6. Attiva e valida il workspace
```bash
curl --fail --silent http://127.0.0.1:8787/health
thothctl --installation deploy/psd/thothii-installation.yaml workspace inspect --workspace psd-clinical --json
```
## 7. Preprocessing (DWH → FK → schema → Evidence)
```bash
thothctl --installation deploy/psd/thothii-installation.yaml workspace preprocess run --workspace psd-clinical --json
```
Se il run si ferma per la revisione delle join (`manual_review_required`):
```bash
# il curatore rivede i candidati e pubblica psd-clinical/schema/annotations.yaml (commit+push),
# poi l'installazione fa pull, e si esegue:
thothctl --installation deploy/psd/thothii-installation.yaml workspace schema accept --workspace psd-clinical --run <run-id> --yes --json
thothctl --installation deploy/psd/thothii-installation.yaml workspace preprocess run --workspace psd-clinical --resume <run-id> --json
```
## 8. Testa una domanda reale
Apri l'app (`http://localhost:5173`), seleziona `psd-clinical` e crea una sessione con una domanda
in linguaggio naturale. Segui le 8 fasi e conferma ai gate fino all'SQL finale.
## Cosa serve ancora da te
1. **URL + credenziali del repo GitHub** del workspace PSD (passo 1).
2. **X-API-Key DWH** (+ CA, se serve) riusabili dall'ambiente attuale (passo 2).
3. Conferma del **provider/modello LLM** di default (`zai/glm-5.2` è già impostato nel descriptor).
4. Conferma dell'**endpoint ping** REST del DWH (`POST /rpc/ping` con `X-API-Key`) per il
`diagnostics.dwh_rest` del descriptor.
Con questi quattro elementi completiamo i passi 2–4 insieme e lanciamo lo smoke live.
- Ristrutturazione del repo PSD nel layout P1.1 + validazione locale.
- Pubblicazione GitHub + deploy key + config operatore completa (bindings/secret/override).
- Avvio stack + attivazione registry + `thothctl inspect` verde.
- Restano solo: **VPN attiva** → `preprocess run` → smoke live di una domanda PSD (P8 L2).