* fix(deploy): pass CADDY_TLS through to caddy container PR #52 added the {$CADDY_TLS:default} substitution to the Caddyfile but forgot to expose CADDY_TLS to the caddy service in docker-compose.yml. Result: Caddyfile substitution falls back to the default (`tls /certs/fullchain.pem /certs/privkey.pem`) regardless of what the operator wrote into .env, and Caddy crash-loops with "open /certs/fullchain.pem: no such file or directory" on any LE / internal deployment. Compose `- CADDY_TLS` (no `=value`) is the bare-form passthrough — Compose reads the value from .env (or the host shell) at up time. No-op when CADDY_TLS is unset (Caddyfile default kicks in), exact behavior preserved for cert-file deployments. Caught by Keboola's first agnes-dev recreate (kids-ai-data-analysis project, agnes-dev.keboola.com) — VM came up with .env containing CADDY_TLS="tls petr@keboola.com" but Caddy ignored it and tried to load the corp PKI cert file. * docs(changelog): document the CADDY_TLS passthrough fix per discipline rule
7.5 KiB
Changelog
All notable changes to Agnes AI Data Analyst.
Format: Keep a Changelog. Versions follow Semantic Versioning, pre-1.0 — public surface (CLI flags, REST endpoints, instance.yaml schema, extract.duckdb contract) may shift between minor versions; breaking changes called out under Changed or Removed with the BREAKING marker.
CalVer image tags (stable-YYYY.MM.N, dev-YYYY.MM.N) are produced for every CI build; semver tags (v0.X.Y) are cut at release boundaries and reference the same commit as a stable-* tag from the same day.
[Unreleased]
Fixed
docker-compose.ymlcaddy service now passesCADDY_TLSthrough to the container (- CADDY_TLSbare-form passthrough). Without it theCaddyfile{$CADDY_TLS:default}substitution always falls back to cert-file mode regardless of what the operator wrote into.env, and Caddy crash-loops on Let's Encrypt / internal-CA deployments. Should have shipped with #52; first attempt was #55, accidentally closed before merging.
Internal
CLAUDE.md— non-negotiable changelog discipline: every PR touching user-visible behavior must updateCHANGELOG.mdunder## [Unreleased]in the same PR.
0.11.0 — 2026-04-26
First tagged semver release. The version = "2.x" strings that appeared in earlier pyproject.toml snapshots were arbitrary placeholders from the initial scaffold and never reflected actual API maturity — resetting to pre-1.0 to signal that things may still shift.
Added — Auth
- Google Workspace groups on
/profile. OAuth callback fetches the signed-in user's group memberships via Cloud Identity (searchTransitiveGroupswith thesecuritylabel — seedocs/auth-groups.mdfor the GCP setup checklist and thesecurity-vs-discussion_forumgotcha). Profile link added to the user dropdown. - Password reset + invite flows for web and admin (
/auth/password/reset,/admin/users/invite). - Personal access tokens (PAT) with separate
:typ=patJWT claim, per-token revoke, last-used IP tracking, "My tokens" + admin "All tokens" UI. - Email magic-link provider (itsdangerous-signed token).
- Optional
SEED_ADMIN_PASSWORDto pre-hash the seed admin (dev convenience).
Added — Deploy
keboola-deploy.ymlworkflow. Tag-triggered alternative torelease.ymlfor shared dev VMs that want explicit "deploy when I tag" semantics. Publishes immutable:keboola-deploy-<tag>+ floating:keboola-deploy-latestalias.- Caddy + Let's Encrypt + corporate-CA TLS.
Caddyfileparametrized via$CADDY_TLSenv var so a single file serves three regimes: cert-file (corp PKI), Let's Encrypt auto-issue, Caddy-internal-CA. URL-driven cert rotation with self-signed fallback (scripts/grpn/agnes-tls-rotate.sh).docker-compose.tls.ymloverlay closes host:8000when Caddy fronts. dev_instancesschema incustomer-instanceTerraform module gains optionaltls_mode+domain(mirrorsprod_instance).infra-v1.6.0tag.- Optional Google OAuth credentials from Secret Manager. Module reads
google-oauth-client-{id,secret}at boot if present; graceful fallback so non-Google deployments aren't affected. LOCAL_DEV_MODE+make local-dev-up/local-dev-downfor one-keystroke local stack with magic-link auth pre-wired.- Per-developer
dev-<prefix>-latestGHCR alias for branches matching<prefix>/<branch>— push-to-deploy on personal dev VMs. /setupweb wizard for first-time instance setup, plus headlessPOST /api/admin/configureandPOST /api/admin/discover-and-register.- Smoke-test job in CI (Docker-in-CI after every release) +
scripts/smoke-test.shfor post-deploy verification.
Added — CLI
- Wheel distribution + auto-update check on startup.
--versionflag,--dry-run+X/Nprogress onda sync, durable sync (atomic writes + manifest hash + retry on transient errors).- gzip on JSON/HTML responses (server-side).
Added — Data
- Remote query engine. Two-phase BigQuery + DuckDB engine for tables too large to sync locally (
--register-bqflag). - Business metrics. Standardized
metric_definitionstable in DuckDB with starter pack importer (da metrics import). /api/healthreturnsversion,channel,commit_sha,image_tag,schema_version.- Custom connector mount support (
connectors/custom/). - OpenAPI snapshot test for breaking-change detection.
Added — Docs / tooling
docs/auth-groups.md,docs/DEPLOYMENT.md,docs/HACKATHON.md,docs/ONBOARDING.mdrunbooks.scripts/debug/probe_google_groups.py— stdlib-only probe for diagnosing Cloud Identity API issues without a deploy cycle.- Schema migration safety tests (idempotency, data preservation, snapshot).
- Pre-migration snapshot of
system.duckdbbefore schema upgrades. - Auto-generated JWT and session secrets with file persistence (
/data/state/.jwt_secret). - Startup banner logging version, channel, and schema version.
Changed
- BREAKING (deployment) — Caddy compose profile renamed
production→tls. Existingdocker compose --profile production up -dinvocations need to switch. - BREAKING (deployment) — Default
Caddyfilemode is now cert-file (tls /certs/fullchain.pem /certs/privkey.pem); for the previous Let's Encrypt auto-issue behaviour setCADDY_TLS=tls <ops-email>in.env. Seedocs/auth-groups.mdandCaddyfileinline docs. - Schema migration v5→v6→v7: adds
users.active,personal_access_tokenstable,personal_access_tokens.last_used_ip. Auto-applied at boot. - Image-level
AGNES_VERSIONnow sourced frompyproject.tomlat build time (no more drift betweenda --versionand the package metadata). - Vendor-agnostic OSS rule codified in
CLAUDE.md— customer-specific names, hostnames, project IDs belong in consumer infra repos, not in this OSS distribution.
Fixed — Security
- Open-redirect guard for backslash in
safe_next_path. SessionMiddleware max_age=3600 + https_only(was browser-session forever, plain-HTTP-OK).- Timezone-aware datetimes in Keboola metadata cache.
- Atomic magic-link token consumption (closes double-use race under concurrent clicks).
- Bootstrap backdoor closed when passwordless seed admin exists.
- urllib3 1.26→2.6.3 (resolves 4 Dependabot security alerts).
- argon2-cffi adopted for password hashing.
- See docs/padak-security.md for the full audit.
Fixed — Other
uvicorn --proxy-headers --forwarded-allow-ips='*'so OAuth callbacks resolve to https when behind a TLS terminator.scripts/grpn/agnes-tls-rotate.shhardened:--max-redirs 0+--proto '=https'on cert fetch, post-fetch PEM validation (rejects HTML error pages from corp portals),ulimit -c 0to suppress coredumps that could leak the unencrypted privkey, POSIX-safe${arr[@]+"${arr[@]}"}array expansion.scripts/tls-fetch.sh— generic URL fetcher (sm://,gs://,https://,file://) with redirect refusal + PEM validation.kbcstoragemoved to optional dep — unblocks urllib3 security updates; primary Keboola path now uses the DuckDB Keboola extension.- Dependencies consolidated into
pyproject.toml(no morerequirements.txt).
Internal
- Test suite expanded to 1357+ tests (4 layers — unit, integration, web smoke, journey).