Add llama.cpp backend, model gateway, URL relay and browser view

Four features, plus a latent bug fix.

llama.cpp backend. Claude Code only ever speaks the Anthropic Messages
API — confirmed empirically by pointing it at a logging server, which
received POST /v1/messages?beta=true. llama-server implements that
natively (verified in its README, alongside --port default 8080), so
this is a plain base-URL backend with no translation shim, the same
shape as Ollama. Its --api-key defaults to none, so the auth token is a
placeholder Claude Code requires and llama-server ignores.

Model alias fix. ANTHROPIC_DEFAULT_HAIKU_MODEL is documented as "also
used for background functionality", and Triple-C set none of the alias
vars. So on every custom-endpoint backend, Claude Code resolved `haiku`
to an Anthropic model id and sent it to a local server that does not
have it — background features failed silently. All four
ANTHROPIC_DEFAULT_{OPUS,SONNET,HAIKU,FABLE}_MODEL vars are now pinned to
the backend's configured model, with an optional Haiku override, and
blanked for Anthropic and Bedrock so those keep Claude Code's defaults.
The deprecated ANTHROPIC_SMALL_FAST_MODEL is never emitted. Existing
Ollama and OpenAI-Compatible containers are recreated once so the new
env reaches them; the snapshot is preserved.

Model gateway. Optional LiteLLM sibling container, off by default,
mirroring stt.rs — this is what makes real OpenAI usable, since
api.openai.com has no /v1/messages. Pinned to v1.96.0 by tag and digest:
the 1.82.7/1.82.8 malware was PyPI-only and never affected the official
images, which is precisely why this builds FROM the image rather than
pip-installing, but 1.84.0 is still the floor for proxy CVEs (API-key
SQLi, Host-header auth bypass, MCP auth bypass). Binds 0.0.0.0 because
project containers consume it, and therefore always sets a master_key —
LiteLLM without one accepts any key. The provider key lives in the OS
keychain and is uploaded into a volume, never an image layer or label.

URL relay. A container-side xdg-open/BROWSER shim opens URLs in the
host's browser. Uses an OSC sequence to /dev/tty rather than a printed
sentinel, because the shim usually runs as a grandchild of a process
capturing its children's output. Degrades to printing the URL when no
terminal is attached, so scheduled tasks do not hang. Only http/https,
with control characters rejected before new URL() — which strips
newlines, so java\nscript: would otherwise parse as javascript:. Nothing
auto-opens; the user confirms. The web terminal shows a tap-to-open
banner instead, since that browser may be a phone across a tunnel.

Browser view. A Project Home tab that watches and takes over the browser
Claude drives with Playwright, using Playwright's own dashboard. Zero
image cost — Playwright stays user-installed. It does not reuse the auth
bridge's PortForward, which binds an unauthenticated port: correct for a
throwaway OAuth listener, wrong for mouse and keyboard control of a
browser in a passwordless-sudo container. Instead a token-gated loopback
proxy checks Host, then token or a forbidden-header origin signal,
before a byte reaches the container. Host ports are confined to
47820..=47827 so CSP frame-src can enumerate them rather than widening
to a wildcard, with a test asserting the two agree.

188 frontend tests, 107 Rust tests, both builds clean.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-08-09 16:55:28 -07:00
co-authored by Claude Opus 5
parent 7d00390e1f
commit cc5f691677
46 changed files with 6194 additions and 61 deletions
+13 -1
View File
@@ -242,6 +242,18 @@ if [ -n "${TZ:-}" ]; then
fi
fi
# ── Browser / URL relay ──────────────────────────────────────────────────────
# Tools that open a browser (gh, aws sso login, gcloud, python webbrowser, ...)
# consult $BROWSER first. Point it at the relay shim, which forwards the URL to
# the host's browser over the terminal. The image already sets this as an ENV —
# that is what `docker exec` terminal sessions inherit — but exporting it here
# means the entrypoint's own children see it too, and (below) that it is
# captured into the cron environment file for scheduled tasks. Under cron there
# is no terminal, so the shim degrades to printing the URL into the task log.
if [ -x /usr/local/bin/triple-c-open ]; then
export BROWSER=/usr/local/bin/triple-c-open
fi
# ── Scheduler setup ─────────────────────────────────────────────────────────
SCHEDULER_DIR="/home/claude/.claude/scheduler"
mkdir -p "$SCHEDULER_DIR/tasks" "$SCHEDULER_DIR/logs" "$SCHEDULER_DIR/notifications"
@@ -255,7 +267,7 @@ ENV_FILE="$SCHEDULER_DIR/.env"
: > "$ENV_FILE"
env | while IFS='=' read -r key value; do
case "$key" in
ANTHROPIC_*|AWS_*|CLAUDE_CODE_*|TRIPLE_C_PERMISSION_MODE|PATH|HOME|LANG|TZ|COLORTERM)
ANTHROPIC_*|AWS_*|CLAUDE_CODE_*|TRIPLE_C_PERMISSION_MODE|PATH|HOME|LANG|TZ|COLORTERM|BROWSER)
# Escape single quotes in value and write as KEY='VALUE'
escaped_value=$(printf '%s' "$value" | sed "s/'/'\\\\''/g")
printf "%s='%s'\n" "$key" "$escaped_value" >> "$ENV_FILE"