chore: back-merge v0.5.0 into develop - #111
Merged
Merged
Conversation
Rewrite README.md and README_zh.md to lead with the TCO-vs-performance
positioning: Ollama-level TCO with vLLM-level throughput, enabled by
putting an AI agent in the operator seat.
- Drop stale MCP tool counts; frame AIMA as an MCP server that plugs
into external agent runtimes (OpenClaw in production), alongside the
internal Explorer PDCA agent.
- Replace per-machine benchmark callouts with aggregate release-gate
numbers: 7 GPU/NPU vendors, 5 OS families, 16 UAT items, 1,200+
evidence files across 86 sub-folders, ~1,000 h on-silicon runtime.
- Rewrite Quick Start as a 5-step guided flow (download, hal detect,
init, onboarding wizard / deploy, OpenAI-compatible call) so new
users have a clear path from binary to first inference.
- Run humanizer/humanizer-zh over both versions to strip em-dash
cascades, bold emphasis, three-part parallelisms, meta labels
("The punchline.", "Why this matters for you.") and self-referential
constructions.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…I GIF (#43) * docs(readme): v2 ship — theatrical sections + visual anchors + β WebUI GIF Direct-to-master per PMM 2026-04-24 — README is user-facing and the updated positioning should land in front of visitors immediately rather than wait for the next release cycle. Changes: - Hero: banner SVG + Orbitron Typing SVG + 4-badge wall - Narrative: four theatrical section titles — 60-Second Genesis / Eight Silicon Ecosystems / The L0→L3 Intelligence Ladder / The Forge - Visuals (new hero images in docs/assets/): architecture-ladder.png 4-layer intelligence ladder eight-silicon-ecosystems.png 8-vendor GPU/NPU matrix supported-engines.png vLLM / llama.cpp / SGLang / Ollama banner.svg brand-coloured repo banner onboarding-webui.gif β WebUI demo — natural-language query → L3a Agent → hardware.detect MCP tool in ~6s - News timeline, collapsible MCP tools breakdown (11 domains, real tool names grepped from internal/mcp/), Forge stat-badge row - Numbers aligned to CLAUDE.md (61 MCP tools, 8 hardware platforms) - GitHub Stars badge hidden until stars ≥ 500 (per PMM marketing principle — low counters on a young repo undercut positioning) README_zh.md delta will follow in a separate PR (§5 of docs/superpowers/specs/2026-04-23-aima-readme-v2-draft.md). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> * docs(readme): address skyguan92 PR review — fix CLI drift + drop Ollama PR #43 review caught six issues (4 inherited from prior README, 2 new in this PR). All addressed in one commit so re-review is straightforward. CLI accuracy (existed in prior README — fix while we're touching it): - `cd aima` → `cd AIMA` (Linux case-sensitive) - `sudo aima init` → `sudo aima onboarding init --tier k3s --yes` (`init` deprecated; current path is onboarding subcommand with explicit tier) - `aima deploy apply --model X` → `aima deploy <model>` (verified against internal/cli/deploy.go on master) - `aima discover` → removed (root cmd doesn't exist; use `aima fleet devices`) - `aima init` references in Web UI / Security sections → `aima onboarding init` Truth-source alignment (newly introduced in v2; reviewer right): - Drop Ollama from `## Supported Engines` — `catalog/engines/` has no ollama.yaml and no Ollama wiring in `internal/`. Was over-claiming. - Text: "four inference runtimes ... and Ollama" → "three" - Image: regenerated supported-engines.png as 1×3 (vLLM / llama.cpp / SGLang) per PMM 2026-04-27, replacing the 2×2 4-engine version - Alt text: Ollama clause removed - Drop `model.safetensors` and `benchmark.ensure_assets` from the MCP tools breakdown. Replaced with skyguan92's canonical lists: model: list/scan/pull/import/info/remove benchmark: run/matrix/record/list Spec source (`docs/superpowers/specs/2026-04-23-aima-readme-v2-draft.md`) also synced — that file is the prompt source for ChatGPT image regen. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
…LAUDE.md ref Follow-up to #43: the §"16 UAT items" section pointed readers at CLAUDE.md / Remote Test Lab for per-device evidence, but that file is internal operator-facing material (SSH hosts, IPs, contact info). Replace with the public evidence chain: docs/uat/v0.4-release-uat.md and artifacts/uat/v0.4/ (u1..u15 — install / hardware detect / model deploy / API / MCP / fleet / onboarding / failover). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
… source (#46) README_zh.md rewrite to match the v2 English README that landed in 3913c42 + c9477c6: - Banner + bilingual switcher (中文 highlighted) + Chinese Typing SVG (Noto Sans SC font, brand purple) - 4 theatrical section headings translated: 60-Second Genesis → 60 秒创世 Eight Silicon Ecosystems → 八大硅基生态 The L0→L3 Intelligence → L0→L3 智能阶梯 The Forge → 熔炉 — 1200 次真机验证 - Numbers and CLI commands aligned to current master (61 MCP tools, 8 ecosystems, `cd AIMA`, `aima onboarding init --tier k3s --yes`, `aima deploy <model>`, `aima fleet devices`) - Three engines (no Ollama), per skyguan92 review - Forge UAT pointer goes to public docs/uat/v0.4-release-uat.md and artifacts/uat/v0.4/, mirroring c9477c6 in English README - Collapsible MCP tools surface — 11 domains, real tool names from internal/mcp/ Also commits docs/superpowers/specs/2026-04-23-aima-readme-v2-draft.md — the design spec used as the prompt source for ChatGPT image regen of architecture-ladder.png / eight-silicon-ecosystems.png / supported-engines.png. PR #43 review noted this file was referenced but missing from master; adding it for traceability so future regeneration runs from a versioned source. Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
…te VRAM
llama.cpp's default n_gpu_layers=999 ("offload every layer") assumes the
model fits in VRAM. On a discrete GPU with less VRAM than the model needs
(e.g. qwen3.5-9b on a 4GB RTX 2050) it forces an impossible allocation and
CUDA-OOMs at load, surfacing to users as a deployment that never becomes
ready.
Cap the engine-declared offload knob (offload_config_key) to CPU when a
discrete GPU can't hold the model's footprint, and warn with the numbers.
Unified-memory hosts (APU/GB10/Apple) share system RAM with the GPU, so
full offload is correct there and is left untouched. The knob is read from
catalog YAML, so no engine name or config key is hardcoded (INV-1/2).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A model pull streamed straight to disk with no free-space check. On a small disk a multi-GB GGUF could fill the filesystem past k3s's disk-pressure eviction threshold, tainting the node NoSchedule so the pod stayed Pending forever — surfacing only as "Timed out waiting for deployment to be ready". Guard the download: the moment its total size is known (reported by the existing downloader progress, before transfer for HF/ModelScope), abort if it won't fit while preserving a disk-pressure reserve (10% of the volume, 2 GiB floor), returning a clear, fatal, pre-transfer error instead of a disk-filling hang. Adds hal.DiskUsage for cross-platform free/total bytes. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
# Conflicts: # README.md # README_zh.md # design/knowledge-yaml-generation-and-consumption-v0.4.md # internal/openclaw/config.go # internal/openclaw/sync_test.go # internal/proxy/sync.go
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Back-merge the audited
v0.5.0release branch after production release.This preserves the release-only metadata, master README v2/assets, and the resolved OpenClaw/proxy history on
develop, as required by the repository release flow.Validation already completed on the identical release tree:
go test ./...go vet ./...go test -race ./...