Local AIPROPRO REQUIREDFC-AI-007
Embed Claw
embed-claw
v0.2.0May 22, 2026Embedding quality — model selection, similarity benchmarks, storage optimization, and retrieval accuracy testing
Embeddings are the silent backbone of every RAG and semantic search pipeline — and most of them are broken. I care about the actual retrieval hit rate at k=5, not the leaderboard MTEB score. I will benchmark your embedding model against YOUR data, not a generic benchmark: measure cluster separation with silhouette scores, find queries where cosine similarity returns noise, and detect modality gaps in multilingual embeddings where "English query → Spanish document" silently fails. I generate embedding-model selection matrices with per-use-case recall curves, not generic ranking tables.
OpenClaw
Hermes
Claude Code
Codex
Cursor
Windsurf
Aider
OllamaPRIMARY ACTION
Unlock with ProCOMPATIBLE WITH
OpenClawHermesClaude CodeCodex+4
8 compatible tools4 output formats
When to Use
- Run local models with private workflows
- Tune inference for local hardware
- Choose effective GGUF variants
- Benchmark practical latency and quality
Compatible Frameworks
OpenClawHermesClaude CodeCodexCursorWindsurfAiderOllama8 TOOLS
Quality Gates
- Embedding models compared for quality
- Vector index appropriate for scale and latency
- Complete search pipeline configured
- Scaling strategy defined with benchmarks
- Quality evaluation included
4 GATES DEFINED
Expected Outputs
Embedding Model Selection Matrix with Per-Use-Case Recall CurvesRetrieval Audit Report with Hit-Rate@K and False-Positive AnalysisMultilingual Gap Scan with Language-Pair Performance HeatmapCluster Separation Report with t-SNE/UMAP Visualisations and Silhouette Scores
Native exports per tool
OpenClaw10 files
openclaw/AGENTS.mdopenclaw/SOUL.mdopenclaw/TOOLS.md+7 moreHermes5 files
hermes/skills/flickclaw/embed-claw/SKILL.mdhermes/skills/flickclaw/embed-claw/references/workflow.mdhermes/skills/flickclaw/embed-claw/references/quality-gates.md+2 moreClaude Code6 files
claude-code/CLAUDE.mdclaude-code/.claude/skills/embed-claw/SKILL.mdclaude-code/.claude/skills/embed-claw/references/workflow.md+3 moreCodex5 files
codex/AGENTS.mdcodex/.flickclaw/agents/embed-claw/codex.mdcodex/.flickclaw/agents/embed-claw/workflow.md+2 moreCursor3 files
cursor/.cursor/rules/flickclaw-embed-claw.mdccursor/.cursor/rules/flickclaw-embed-claw-workflow.mdccursor/.cursor/rules/flickclaw-embed-claw-quality-gates.mdcWindsurf3 files
windsurf/.windsurf/rules/flickclaw-embed-claw.mdwindsurf/.windsurf/rules/flickclaw-embed-claw-workflow.mdwindsurf/.windsurf/rules/flickclaw-embed-claw-quality-gates.mdAider3 files
aider/CONVENTIONS.mdaider/aider.mdaider/.aider.conf.ymlOllama4 files
ollama/Modelfileollama/system-prompt.mdollama/template.md+1 moreInstall Commands
Install the FlickClaw CLI, then select your AI agent framework below to get the correct install command.
Step 1: Install CLI (one-time)
npm install -g @flickclaw/cli@latestStep 2: Select Framework
OpenClaw
npm exec --yes @flickclaw/cli@latest -- install embed-claw --target openclawDownload as ZIP
Example Prompt
Try this prompt with Embed Claw to see what it can do:
Optimize the local model setup for performance. Benchmark current config and suggest improvements for Embedding Model Selection Matrix with Per-Use-Case Recall Curves, Retrieval Audit Report with Hit-Rate@K and False-Positive Analysis, Multilingual Gap Scan with Language-Pair Performance Heatmap.Example Output
IllustrativeWhat a typical Embed Claw report looks like:
# Embed Claw — Assessment Report **Project**: ollama-deployment **Context**: a local LLM deployment running Llama 3.1 8B on consumer GPU hardware **Generated**: 2026-07-10 ## Executive Summary The Embed Claw completed its review of ollama-deployment (a local LLM deployment running Llama 3.1 8B on consumer GPU hardware). 3 findings were identified with concrete remediation steps. All quality gates were verified before delivery. ## Findings | # | Severity | Area | Finding | Recommended Action | |---|----------|------|---------|-------------------| | 1 | **P1** | Performance | Inference latency spikes to 8s under concurrency | Enable continuous batching and set max_batch=4 | | 2 | **P2** | Memory | Model consumes 18GB VRAM, headroom insufficient | Switch to Q4_K_M quantization, target <12GB | | 3 | **P2** | Setup | No Modelfile with system prompt defined | Create Modelfile with role, constraints, and templates | ## Quality Gates - [✓] retrieval_hit_rate_at_k5_on_your_data - [✓] cluster_separation_silhouette_score - [✓] multilingual_query_document_gap_scan ## Outputs Generated - **Embedding Model Selection Matrix with Per-Use-Case Recall Curves**: Included in the report above. - **Retrieval Audit Report with Hit-Rate@K and False-Positive Analysis**: Included in the report above. - **Multilingual Gap Scan with Language-Pair Performance Heatmap**: Included in the report above. - **Cluster Separation Report with t-SNE/UMAP Visualisations and Silhouette Scores**: Included in the report above. ## Validation - [x] All quality gates passed (3/3) - [x] 3 findings documented with severity and remediation - [x] 4 output sections generated - [x] Evidence collected and referenced --- *This is an illustrative example output from FlickClaw. Results vary based on project context.*
RELATED AGENTS