fix: stabilize onboarding and thread selection e2e

.env.example

@@ -91,10 +91,11 @@ OPENHUMAN_MODEL=
 # [optional] Workspace directory (default: ~/.openhuman or ~/.openhuman-staging when OPENHUMAN_APP_ENV=staging)
 # Leave commented out to use the default. Setting to empty string is treated as unset.
 # OPENHUMAN_WORKSPACE=
-# [optional] Master key for the `encrypted_file` keyring used in staging/production,
-# as 64 hex characters (`openssl rand -hex 32`). Headless/container deploys have no
-# OS keychain to hold it; set exactly one of these two, injected from your secret
-# manager. Losing the key orphans every secret encrypted under it.
+# [required without an OS keychain] Master key for the `encrypted_file` keyring,
+# which is the default unless OPENHUMAN_APP_ENV=dev/development or an explicit
+# backend override is set. Use 64 hex characters (`openssl rand -hex 32`).
+# Headless/container deploys should set exactly one of these two through a
+# secret manager. Losing the key orphans every secret encrypted under it.
 # OPENHUMAN_KEYRING_MASTER_KEY=
 # OPENHUMAN_KEYRING_MASTER_KEY_FILE=/run/secrets/openhuman_master_key
 # [optional] Default: 0.7

.github/ci-paths-filter.yml

@@ -44,6 +44,20 @@ docs:
   - 'scripts/generate-architecture-docs.mjs'
   - 'scripts/__tests__/generate-architecture-docs.test.mjs'
   - 'gitbooks/developing/architecture/frontend.md'
+  - 'gitbooks/developing/embed/**'
+  - 'gitbooks/developing/embedding.md'
+  - 'gitbooks/SUMMARY.md'
+  - 'docs/gitbooks/en/developing/embed/**'
+  - 'docs/gitbooks/en/developing/embedding.md'
+  - 'docs/gitbooks/en/SUMMARY.md'
+  - 'crates/openhuman-embed/**'
+  - 'scripts/generate-embed-docs.mjs'
+  - 'scripts/run-embed-examples.mjs'
+  - 'scripts/__tests__/generate-embed-docs.test.mjs'
+  - 'scripts/__tests__/run-embed-examples.test.mjs'
+  - 'llms.txt'
+  - 'llms-full.txt'
+  - 'EMBED.md'
   - 'package.json'
 rust-core:
   - '.github/workflows/ci-*.yml'

.github/workflows/test-reusable.yml

@@ -244,6 +244,9 @@ jobs:
           # The core binary and developer bins live in the CLI crate.
           bash scripts/ci-cancel-aware.sh cargo test -p openhuman-cli --bins --features "${CLI_FEATURES}"
           bash scripts/ci-cancel-aware.sh cargo test -p openhuman --doc --features "${FEATURES}"
+          bash scripts/ci-cancel-aware.sh cargo test -p openhuman-embed --doc --features "${FEATURES}"
+          RUSTDOCFLAGS="-D warnings" bash scripts/ci-cancel-aware.sh cargo doc -p openhuman-embed --no-deps --features "${FEATURES}"
+          node scripts/run-embed-examples.mjs
 
           while IFS= read -r target; do
             [ -n "${target}" ] || continue

.gitignore

@@ -156,6 +156,3 @@ diff-coverage.md
 ci-out/
 ci-out-ex63/
 
-# Cargo build output inside the in-tree motosan-ai-oauth crate
-vendor/motosan-ai-oauth/target/
-

AGENTS.md

@@ -405,6 +405,12 @@ Additional rules:
 
 ## Tool, harness, and runtime boundaries
 
+Embed public API documentation: [Embedding](gitbooks/developing/embed/README.md),
+[concepts](gitbooks/developing/embed/concepts/README.md), and the source-generated
+[builder setters](gitbooks/developing/embed/builder-setters.md). Snippets come from
+compiled examples; run `pnpm docs:generate` and `pnpm docs:check` after changing them.
+
+
 `tinyagents` owns tool-call dialects, parsing, catalog rendering, transcript
 replay, session identity, and the agent loop. `tinytools` owns the shared
 `Tool` trait and tool types. OpenHuman owns execution policy, approvals,
@@ -553,7 +559,6 @@ Direct rendered submodules under `vendor/`:
 | `tinyskills` | Host-independent skill/workflow bundle parsing, discovery, scope resolution, resource inventory, and safe reads. OpenHuman owns trust and execution policy. |
 | `tinyvoice` | Host-agnostic voice primitives such as audio framing, VAD, wake-word gating, routing, and STT hallucination detection. |
 | `tinywallet` | Multi-chain wallet: `tinywallet-crypto` (address, asset, chain, `rpc::Transport`, tx codec), `tinywallet-x402` (x402 wire, payment, spending ledger, `x402_request` tool), `tinywallet-web3` (wallet engine, per-chain build/sign/broadcast flows, swap/bridge/dapp quotes, agent tools) behind host seams (`WalletSigner`, `PaymentSigner`, `WalletAccounts`, `RpcEndpoints`, `QuoteScope`, `Web3Backend`, `ProxyPolicy`), and the loadable `tinywallet-module` that derives keys and signs. OpenHuman keeps keyring, consent, credentials, config, controllers and the seam impls under `web3/`. |
-| `motosan-ai-oauth` | Provider-agnostic PKCE OAuth login and token-refresh primitives. |
 
 Some rendered submodules are shared dependencies nested inside those projects,
 not separate OpenHuman feature implementations. Make changes to them in their