LYGO-native local second brain (Obsidian vault + Ollama). Ingest raw sources, embed index, search, multi-model consensus, wiki pages from vault only. Use for...
安全
LYGO pxpipe-LYGO
试用Multi-tool vision context compression for LYGO (Anthropic, OpenAI, Grok/xAI, Gemini). Use before stuffing huge prompts or logs into pay-to-go APIs. Requires...
它能做什么
Multi-tool vision context compression for LYGO (Anthropic, OpenAI, Grok/xAI, Gemini). Use before stuffing huge prompts or logs into pay-to-go APIs. Requires lygo-protocol-stack; read references/SECURITY.md. Agents run pxpipe_lygo_for_agent.py or local proxy — no auto publish/git push.
技能文档
LYGO pxpipe-LYGO (ClawHub)
Sovereign pxpipe-style compression: bulky text → PNG for vision APIs, P0 gate, manifests under data/pxpipe_lygo/.
When to use
- Token saver on large context (≥800 chars), not byte-exact secrets/hashes.
- Before pasting multi-KB tool dumps into Grok, Claude, or GPT.
When not to use
- API keys, seeds, Merkle hashes, or line-precise diffs you must reproduce exactly.
- Unknown
LYGO_STACK_ROOTor without readingreferences/SECURITY.md.
Setup
npx clawhub@latest install deepseekoracle/lygo-pxpipe-lygo
export LYGO_STACK_ROOT=/absolute/path/to/lygo-protocol-stack
pip install -r "$LYGO_STACK_ROOT/requirements-pxpipe.txt"
Agent commands (no proxy)
cd "$LYGO_STACK_ROOT"
python tools/pxpipe_lygo_for_agent.py --shrink-file path/to/huge.txt --target grok
python tools/pxpipe_lygo_for_agent.py --file path/to/huge.txt --target anthropic --png .pxpipe_ctx.png
Multi-tool proxy
python tools/run_pxpipe_lygo_proxy.py
| Client | Base URL |
|---|---|
| Anthropic / Claude Code | http://127.0.0.1:47821 |
| OpenAI SDK | http://127.0.0.1:47821/v1 |
| Grok / xAI | http://127.0.0.1:47821/v1 |
Set the matching API key in the environment. Proxy forwards upstream; does not store keys in manifests.
Agent rules
- Read
references/SECURITY.mdon first use. - Never compress secrets or
.envcontents. - No
git push, ClawHub publish, or HF upload unless user explicitly asks. - Prefer
--shrink-filein chat threads; use full JSON blocks only when building API calls.
Skill chain
lygo-api-token-saver → lygo-pxpipe-lygo → lygo-protocol-stack-operator
Doc
docs/BIOPHASE7_PXPIPE_LYGO.md in the stack repo.
相关技能
LYGO Prompt Implant System (LPIS) v1.1 — analyze authorized prompts locally; P0 gate, P1 vault, P3 sovereign variants, P5 advisory implant. Requires --i-auth...
Minimize pay-to-go xAI/API token spend. Use when user says token saver, save tokens, API budget, pay-to-go burn, or before long Grok/Claude/GPT calls. Prefer...
LYGO Context Guard — pre-flight token budget, secret redaction, and deterministic context compaction for AI agents. Use when stuffing tool dumps, logs, files, or long chats into a model; when context is too large; when you need to save tokens / reduce API cost; when you must redact API keys before re-injection. Pure local stdlib. No network, no subprocess. Commands: estimate, redact, compact, budget, toolpack, preflight. Install clawhub:@deepseekoracle/lygo-context-guard.
Give your agent its own LLM gateway account. Signs up for ProxyLLM (an OpenAI-compatible gateway with fallback lanes, budgets, and request logs) autonomously...
LYGO SMART DISK AGENT — lean 100% LYGO CLAW. Localhost portal with local operator token (not cloud password), P0–P5, host Ollama. No HTTP chat-memory export....