3 Commits

Author	SHA1	Message	Date
Teknium	4d1880e0bf	fix(integration): restore subprocess encoding/stdin guards dropped in simplification ... Simplification workers collapsed subprocess call sites into shared kwargs helpers and dropped the Windows/TUI safety kwargs on the way: - encoding='utf-8', errors='replace' restored on text=True runs in copilot_acp_client, hermes_cli/setup (vercel install), managed_uv (codesign steps), local_runtime/hardware._stdout, a2a adapter. - stdin=subprocess.DEVNULL restored on copilot probe, verify/runner _SUBPROCESS_KW, iron_proxy._run, google_meet playwright/system_profiler, simplex convert, whatsapp _RUN_TEXT, mem0 ollama serve Popen. google_meet sudo/brew install keeps inherited stdin (user-confirmed, may prompt) — marked noqa: subprocess-stdin. - Windows-safe SIGKILL: getattr(signal, 'SIGKILL', SIGTERM) in verify/runner; photon _kill call re-marked windows-footgun: ok (unreachable on win32). - scripts/check_subprocess_stdin.py now recognizes **kwargs splats (**_KW / **_kw(...)) ONLY when the same-file definition provably sets stdin= — covers tui_gateway _capture_run_kwargs/run_kw. Parity test added.	2026-09-02 16:36:10 -07:00
Teknium	3ffd44acd3	refactor(hclib): models/runtime — models, inventory, runtime_provider, provider catalog, local_runtime, banner	2026-09-02 14:44:48 -07:00
emozilla	43e67d872f	feat: local models — managed llama.cpp runtime with one-click desktop setup ... Run models locally as a first-class provider. The CLI grows a managed llama.cpp runtime (engine install, model download, server supervision); the desktop app grows the full setup and management story on top of it. GUI surfaces ship behind the desktop --local launch flag (hermes desktop --local, or the flag on the packaged app); backend routes and the CLI are always live. Runtime (hermes_cli/local_runtime/): - curated GGUF catalog with per-machine variant selection: hardware probe (VRAM/RAM/UMA), fit planning with spill accounting, quant choice by context window - derived recommendation: quality-ranked picks gated by a predicted decode-speed floor, bandwidth-aware on unified memory; the decision table is pinned as a test (pick AND reason per memory class), and the Recommended badge explains its pick in a tooltip fed by the resolver's actual branch - engine install + model download with resumable split parts, cumulative plan-level progress, and staged-model integrity (a split GGUF counts only when every part is present) - server supervision: spawn/adopt/stop, router mode with per-model load progress relayed over SSE, abandoned-request cleanup Desktop: - Settings -> Providers -> Local models: one-click quickstart (install engine, download the recommended model, boot) plus per-model download/ activate/eject, fit-ranked catalog with context pills - model pickers (composer dropdown + Cmd+K) show staged local models, in-flight downloads as live progress rows, and load-into-memory bars - local-setup campaign tip for eligible hardware; System resources statusbar widget (GPU/VRAM/RAM); in-chat load progress during sends - friendly dead-server errors, and failed agent builds retry on the next send instead of wedging the session Co-developed with NVIDIA field feedback on RTX 5090 and DGX Spark.	2026-09-01 16:01:53 -04:00