AGENT SKILL

setup-and-gateway

Get OpenSquilla installed, onboarded, and running through a ready local gateway and Web UI.

License 许可
Apache-2.0
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
setup-and-gateway
license
Apache 2.0
metadata
{ "disco-role": "operating" }
description
Get OpenSquilla installed, onboarded, and running through a ready local gateway and Web UI.
disable-model-invocation
true

02 资源文件 4

03 SKILL.md 正文

Setup and Gateway

Use this sub-skill for the path from a fresh install to the first successful gateway and Web UI run.

Route elsewhere

  • Provider catalog, router tier selection, and search provider setup: ../configuration-and-routing/SKILL.md
  • CLI automation, sessions, memory, diagnostics, replay, migration, and uninstall: ../cli-and-automation/SKILL.md
  • Terminal UI and desktop runtime details: ../tui-and-desktop/SKILL.md

What belongs here

  • release-wheel vs source-install choices
  • first-run onboarding and safe reruns with --if-needed or --minimal
  • gateway lifecycle: run, start, status, stop, restart, reload
  • doctor readiness checks and bundle diagnostics collection
  • Web UI launch basics and local access
  • source-install prerequisites and built-console expectations
  • gateway bind/port behavior and safe exposure notes

Fast operating path

  1. Pick the install path.
    • Release wheel: normal terminal install, no Git or Node.js required.
    • Source install: only when the user is working from a checkout and needs to build the Web UI.
  2. Run onboarding.
    • Interactive: opensquilla onboard
    • Idempotent rerun: opensquilla onboard --if-needed
    • Smaller first-run setup: opensquilla onboard --minimal
    • Script-friendly provider setup: opensquilla onboard --provider <id> --api-key-env <ENV_VAR>
  3. Start the gateway.
    • Foreground: opensquilla gateway run
    • Managed background process: opensquilla gateway start --json
  4. Confirm readiness.
    • opensquilla onboard status
    • opensquilla gateway status
    • opensquilla doctor
    • Open http://127.0.0.1:18791/control/
  5. If the user needs public exposure, use the safer bind and auth notes in references/gateway-lifecycle.md first.
  6. Use references/troubleshooting.md for the common first-run failures.

When to hand off

  • If the question is really about provider, router, or search setup, hand off to configuration-and-routing.
  • If the question is about chat history, sessions, memory, cron, replay, or uninstall, hand off to cli-and-automation.
  • If the question is about TUI or desktop packaging/runtime behavior, hand off to tui-and-desktop.

Bundled references

  • references/install-and-first-run.md
  • references/gateway-lifecycle.md
  • references/web-ui.md
  • references/troubleshooting.md

Notes

  • opensquilla bundle is a shareable diagnostics zip and works even when the gateway will not start.
  • Source installer scripts are reference-only here because they mutate the install and rebuild the Web UI; this sub-skill stays guidance-only.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351