mirror of
https://github.com/rennf93/roboco.git
synced 2026-08-03 07:23:24 +02:00
feat(grok-cli): grok CLI agent image + headless entrypoint
The roboco-agent-grok image now installs xAI's official grok CLI (Grok Build, pinned 0.2.56) instead of opencode, authenticated by the SuperGrok subscription via a mounted ~/.grok/auth.json (parity with the Claude ~/.claude mount, no metered API key). The entrypoint renders ~/.grok/config.toml + per-role flags from /app (the ModuleNotFound-shadowing lesson), runs grok -p headless with --output-format json, keeps the prompt-injection guard, and exits 75 on a rate-limit so the orchestrator parks the provider. No in-container SDK server or budget-feed — native --max-turns + server-side terminal-substitute replace them.
This commit is contained in:
Renn F
@@ -1,55 +1,41 @@
|
||||
# Grok (xAI) Agent Image
|
||||
# =============================================================================
|
||||
# Runs grok-build-0.1 through the opencode CLI (OpenAI protocol) instead of
|
||||
# Claude Code, while reusing the base image's roboco venv + uv + the RoboCo MCP
|
||||
# gateway servers. The entrypoint renders opencode.json from the spawn env +
|
||||
# mounted mcp-config.json (see roboco.llm.providers.opencode_config) and runs
|
||||
# opencode. One runtime image serves every role — role behaviour comes from the
|
||||
# mounted system prompt / manifest / mcp-config, exactly as on the Claude path.
|
||||
# Runs Grok Build through xAI's official `grok` CLI, authenticated by the
|
||||
# SuperGrok subscription via a mounted ~/.grok/auth.json — the parity analogue of
|
||||
# the Claude Code path's mounted ~/.claude (no metered API key). Reuses the base
|
||||
# image's roboco venv + uv + the RoboCo MCP gateway servers. The entrypoint
|
||||
# renders ~/.grok/config.toml (the gateway) + the per-role grok flags from the
|
||||
# mounted mcp-config.json (see roboco.llm.providers.grok_cli_config) and runs the
|
||||
# CLI headless. One runtime image serves every role — role behaviour comes from
|
||||
# the mounted system prompt / manifest / mcp-config, exactly as on the Claude path.
|
||||
# =============================================================================
|
||||
|
||||
FROM roboco-agent-base
|
||||
|
||||
USER root
|
||||
|
||||
# opencode — the OpenAI-protocol agent runtime. grok-build-0.1 runs on opencode's
|
||||
# BUILT-IN xai provider (no custom provider block / npm needed), so only
|
||||
# opencode-ai is installed; it resolves the provider SDK at runtime.
|
||||
# Pinned: this version is the live-verified runtime (grok-build-0.1 on the NAS).
|
||||
# Untrusted model output runs under it, so bump the pin deliberately, never float.
|
||||
ARG OPENCODE_VERSION=1.17.8
|
||||
RUN npm install -g "opencode-ai@${OPENCODE_VERSION}" \
|
||||
&& npm cache clean --force \
|
||||
&& rm -rf /root/.npm /tmp/*
|
||||
# Install the official grok CLI (Grok Build) for the agent user: the binary lands
|
||||
# at ~/.local/bin/grok and its runtime (bin / bundled / skills) at ~/.grok, both
|
||||
# agent-owned. Pinned — untrusted model output runs under it, so bump the version
|
||||
# deliberately, never float. (curl + bash are provided by roboco-agent-base.)
|
||||
ARG GROK_CLI_VERSION=0.2.56
|
||||
RUN su agent -s /bin/bash -c "export HOME=/home/agent; \
|
||||
curl -fsSL https://x.ai/cli/install.sh | bash -s ${GROK_CLI_VERSION}" \
|
||||
&& rm -rf /tmp/*
|
||||
|
||||
# opencode plugins, baked into the AUTO-DISCOVERY dir (~/.config/opencode/plugin/)
|
||||
# rather than referenced by a config `plugin:` path — the dir is the simplest
|
||||
# registration route. Each plugin uses a NAMED export (opencode's convention).
|
||||
# secret-scrub — bash-guard parity (PAT/credential deny on tool.execute.before)
|
||||
# budget-feed — POSTs budget/loop/terminal counters to the in-container SDK
|
||||
# server (tool.execute.{before,after}); the entrypoint starts
|
||||
# that server (roboco.agent_sdk.server) for Claude-parity.
|
||||
COPY docker/grok/secret-scrub.js /home/agent/.config/opencode/plugin/secret-scrub.js
|
||||
COPY docker/grok/budget-feed.js /home/agent/.config/opencode/plugin/budget-feed.js
|
||||
|
||||
# Entrypoint: render opencode.json, then run opencode (overrides base's `claude`).
|
||||
COPY docker/scripts/grok-agent-entrypoint.sh /app/scripts/grok-agent-entrypoint.sh
|
||||
RUN chmod 0755 /app/scripts/grok-agent-entrypoint.sh
|
||||
|
||||
# opencode persists data under ~/.local/share and state under ~/.local/state, and
|
||||
# reads config + plugins from ~/.config/opencode. When the orchestrator
|
||||
# bind-mounts the opencode store at ~/.local/share/opencode, docker creates the
|
||||
# intermediate ~/.local AS ROOT, so the non-root agent can no longer create its
|
||||
# siblings and opencode EACCESes at boot. Pre-create the trees agent-owned so the
|
||||
# mount leaves the parents writable (complements the orchestrator's 0777
|
||||
# host-source pre-create), and so the baked plugin dir is agent-owned.
|
||||
RUN mkdir -p /home/agent/.local/share/opencode /home/agent/.local/state \
|
||||
/home/agent/.config/opencode/plugin \
|
||||
&& chown -R agent:agent /home/agent/.local /home/agent/.config
|
||||
# Entrypoint: render ~/.grok/config.toml + the per-role flags, then run grok
|
||||
# headless (overrides the base image's `claude` entrypoint).
|
||||
COPY docker/scripts/grok-cli-agent-entrypoint.sh /app/scripts/grok-cli-agent-entrypoint.sh
|
||||
RUN chmod 0755 /app/scripts/grok-cli-agent-entrypoint.sh \
|
||||
&& chown -R agent:agent /home/agent/.grok /home/agent/.local
|
||||
|
||||
USER agent
|
||||
|
||||
LABEL role="grok-runtime"
|
||||
LABEL description="Grok (xAI) agent runtime — grok-build-0.1 via opencode (OpenAI protocol)"
|
||||
# grok installs to ~/.local/bin; put it ahead of the venv on PATH so the
|
||||
# entrypoint finds `grok` (and still resolves `python` to /app/.venv/bin).
|
||||
ENV PATH="/home/agent/.local/bin:/app/.venv/bin:$PATH"
|
||||
|
||||
ENTRYPOINT ["/app/scripts/grok-agent-entrypoint.sh"]
|
||||
LABEL role="grok-cli-runtime"
|
||||
LABEL description="Grok (xAI) agent runtime — Grok Build via the official grok CLI"
|
||||
|
||||
ENTRYPOINT ["/app/scripts/grok-cli-agent-entrypoint.sh"]
|
||||
|
||||
Executable
+67
@@ -0,0 +1,67 @@
|
||||
#!/usr/bin/env bash
|
||||
# Entrypoint for the roboco-agent-grok-cli image (one-shot delivery roles).
|
||||
#
|
||||
# Runs an agent on xAI's official `grok` CLI (Grok Build), authenticated by the
|
||||
# SuperGrok subscription via a mounted ~/.grok/auth.json — the parity analogue of
|
||||
# the Claude Code path's mounted ~/.claude. The gateway, identity, and workspace
|
||||
# are mounted by the orchestrator's shared container assembly (the same that
|
||||
# wires Claude); this entrypoint renders the grok runtime config from that mount
|
||||
# and runs the CLI headless.
|
||||
set -euo pipefail
|
||||
|
||||
# Render ~/.grok/config.toml (the MCP gateway) + the per-role grok flags. Run
|
||||
# from /app so `python -m` resolves the INSTALLED roboco package: dev/doc/qa
|
||||
# agents run at their workspace-clone cwd, whose own roboco/ dir would shadow it
|
||||
# on the sys.path front (the ModuleNotFound lesson). The render reads
|
||||
# ROBOCO_MCP_CONFIG + ROBOCO_AGENT_ID and writes the config + an args file.
|
||||
( cd /app && python -m roboco.llm.providers.grok_cli_config )
|
||||
|
||||
GROK_ARGS_FILE="${ROBOCO_GROK_ARGS_FILE:-/tmp/roboco-grok-args}"
|
||||
mapfile -t GROK_ARGS < "$GROK_ARGS_FILE"
|
||||
|
||||
# Prompt-injection guard (parity with the Claude UserPromptSubmit hook): the task
|
||||
# prompt is DATA, not instructions — refuse a poisoned one before the model sees
|
||||
# it. Same patterns as docker/scripts/user-prompt-hook.sh; run from /app too.
|
||||
if ! ( cd /app && python -m roboco.agent_sdk.prompt_guard "${ROBOCO_INITIAL_PROMPT:-}" ); then
|
||||
echo "Refusing to run: task prompt matched a prompt-injection pattern." >&2
|
||||
exit 1
|
||||
fi
|
||||
|
||||
# Run the agent. The prompt comes from an env var (never an untrusted argv
|
||||
# positional). `< /dev/null` keeps the headless run from blocking on stdin. We
|
||||
# do NOT `exec`: the script regains control to inspect the result + exit code.
|
||||
# `--cwd` is the agent's workspace (the orchestrator sets the container workdir
|
||||
# to it, mirroring the Claude path). Per-role flags (tool removal / deny rules /
|
||||
# effort / turn cap) come from the rendered args file.
|
||||
RUN_LOG="/tmp/grok-run.json"
|
||||
ERR_LOG="/tmp/grok-run.err"
|
||||
WORKSPACE="${ROBOCO_WORKSPACE:-$PWD}"
|
||||
set +e
|
||||
grok -p "${ROBOCO_INITIAL_PROMPT:-}" \
|
||||
-m "${ROBOCO_AGENT_MODEL:-grok-build}" \
|
||||
--cwd "$WORKSPACE" \
|
||||
--output-format json \
|
||||
"${GROK_ARGS[@]}" \
|
||||
< /dev/null > "$RUN_LOG" 2> "$ERR_LOG"
|
||||
run_rc=$?
|
||||
set -e
|
||||
# Surface the run output + any stderr into the agent log.
|
||||
cat "$RUN_LOG"
|
||||
[ -s "$ERR_LOG" ] && cat "$ERR_LOG" >&2
|
||||
|
||||
# Rate-limit detection (parity with the opencode B4 path): an xAI 429 / quota
|
||||
# error ends the run without a terminal verb. Detect it from the run output and
|
||||
# exit 75 (EX_TEMPFAIL) so the orchestrator PARKS the grok provider instead of
|
||||
# the dispatcher respawning the same task every tick (429 -> exit -> respawn, a
|
||||
# token loop). A rate-limited task is retried once the limit lifts, not dropped.
|
||||
if grep -qiE '(\b429\b|rate.?limit|too many requests|quota|insufficient_quota)' \
|
||||
"$RUN_LOG" "$ERR_LOG" 2>/dev/null; then
|
||||
echo "[grok] rate-limited — exiting 75 so the orchestrator parks the provider;" \
|
||||
"the task is retried when the limit lifts." >&2
|
||||
exit 75
|
||||
fi
|
||||
|
||||
# A graceful exit without a terminal verb is handled server-side by the
|
||||
# orchestrator (_handle_stopped_container substitutes the still-owned task) —
|
||||
# the grok-cli runtime needs no in-container SDK server for that.
|
||||
exit "$run_rc"
|
||||
Reference in New Issue
Block a user