Skip to main content

Hologres Agent Skills

An interactive installer that distributes Hologres AI agent skills to various AI coding tools.

Included Skills

Skill Depends on Description
hologres-cli — Teaches AI agents how to use the Hologres CLI tool — command usage, safety features, output formats, and best practices
hologres-query-optimizer hologres-cli Enables AI agents to analyze and optimize Hologres SQL query execution plans
hologres-slow-query-analysis hologres-cli Equips AI agents to diagnose slow/failed queries using hologres.hg_query_log
hologres-schema-generator hologres-cli Hologres DDL schema design and table creation expert — storage format selection, index configuration, partition design
hologres-privileges hologres-cli Hologres privilege management using PostgreSQL standard GRANT/REVOKE authorization model
hologres-uv-compute hologres-cli Real-time UV/PV computation pipelines using Dynamic Tables and RoaringBitmap
hologres-bsi-profile-analysis hologres-cli BSI (Bit-Sliced Index) based user profile analysis — tag computation, crowd targeting, GMV analysis
hologres-ad-campaign hologres-cli AI-powered ad creative generation and campaign analysis — video synthesis, virtual delivery simulation, real-time ROI via Dynamic Tables
hologres-instance-health-analyse hologres-cli Instance health diagnosis — Warehouse resource inspection, FAILED query classification, CPU/memory slow query analysis, structured diagnostic reports
hologres-diagnosis-cpu hologres-cli CPU anomaly diagnosis — 4-quadrant root-cause analysis (macro qualitative / Worker-Shard distribution / query attribution / background task interference), structured diagnostic report and governance action list
hologres-diagnosis-memory hologres-cli Memory anomaly diagnosis — OOM / sustained-high / Worker imbalance / leak root-cause attribution along 4 lines (Query / skew / Write & background / System & metadata), with structured Markdown diagnostic report and action checklist
hologres-daily-report hologres-cli Daily ops diagnosis report — opinionated "diagnostic conclusion + root cause + action recommendation" report across 6 dimensions (health, availability, compute, SQL perf, cost governance, capacity forecast)
hologres-knowledge-base hologres-cli Search & RAG knowledge base — full-text inverted index (Tantivy + BM25), HGraph vector index, hybrid retrieval (vector + BM25 + scalar) in one SQL, optional embedding via Hologres ai_gen() or holo-search-sdk

Note: All skills except hologres-cli depend on it as the foundational skill. SQL execution, GUC management, and data operations are performed through CLI commands. Install hologres-cli skill first.

Supported AI Tools

Claude Code, OpenClaw, Cursor, Codex, OpenCode, GitHub Copilot, Qoder, Trae

Quick Start

Install via uvx (Recommended)

# Run directly without installation
uvx hologres-agent-skills

Install via pip

pip install hologres-agent-skills
hologres-agent-skills

Install from source (Development)

cd agent-skills
uv sync
uv run hologres-agent-skills

Usage

The installer guides you through an interactive workflow:

  1. Select tool — Choose which AI coding tool to install skills for
  2. Confirm path — Verify the installation directory
  3. Select skills — Pick one or more skills to install
  4. Done — Skills are copied to the tool's skills directory
$ hologres-agent-skills

🚀 Hologres Agent Skills Installer
==================================================

📋 Select tool to install to:
? Select one tool: Claude Code

📁 Project root: /path/to/your/project
   (Skills will be installed under .claude/skills)
? Install skills to this directory? Yes

📦 Select skills to install:
? Select skills:
  ● hologres-cli
  ● hologres-query-optimizer
  ● hologres-slow-query-analysis
  ● hologres-schema-generator
  ● hologres-privileges
  ● hologres-uv-compute
  ● hologres-bsi-profile-analysis
  ● hologres-ad-campaign
  ● hologres-instance-health-analyse
  ● hologres-diagnosis-cpu
  ● hologres-diagnosis-memory
  ● hologres-daily-report
  ● hologres-knowledge-base

✨ Installation complete

Development

Build & Publish to PyPI

cd agent-skills

# Build only (artifacts in dist/)
python upload_to_pypi.py --build

# Upload to TestPyPI (verification)
export TEST_PYPI_TOKEN="pypi-xxx"
python upload_to_pypi.py --test

# Upload to official PyPI
export UV_PUBLISH_TOKEN="pypi-xxx"
python upload_to_pypi.py --publish

# Bump version and publish
python upload_to_pypi.py --publish --version 0.2.0

Publish to ClawHub (OpenClaw Skill Marketplace)

Publish skills to the ClawHub public skill registry:

cd agent-skills

# Prerequisites
npm i -g clawhub
clawhub login

# Publish all skills
python publish_to_clawhub.py

# Publish a specific skill
python publish_to_clawhub.py --skill hologres-cli

# Dry-run (preview without publishing)
python publish_to_clawhub.py --dry-run

# Bump patch version before publishing
python publish_to_clawhub.py --bump

# Set a specific version
python publish_to_clawhub.py --version 1.2.0

# Publish under an org handle
python publish_to_clawhub.py --owner holomcp

Publish to Aone (contextlab)

Publish individual skills to the Aone platform:

cd agent-skills

# Set authentication token
export AONE_TOKEN=<your-token>

# Publish all skills
python publish_to_aone.py

# Publish a specific skill
python publish_to_aone.py --skill hologres-cli

# Dry-run (preview without publishing)
python publish_to_aone.py --dry-run

# Bump patch version before publishing
python publish_to_aone.py --bump

# Set a specific version
python publish_to_aone.py --version 1.2.0

Catalog Consistency Check

Skill names are duplicated in 5 catalog files (installer + 4 READMEs). Every time a new skill is added under skills/, all 5 must be updated, or uvx hologres-agent-skills users won't see / be able to install it. The check script catches the drift:

# Standalone run (exit 1 on drift, exit 0 on clean)
python agent-skills/tests/test_catalog_consistency.py

# Or via pytest
pytest agent-skills/tests/test_catalog_consistency.py -v

Install as a local pre-commit hook (optional, recommended)

A repo-local pre-commit hook lives at .githooks/pre-commit. It chains the user's globally-configured pre-commit (e.g. Alibaba AccessKey scanner) and then runs the catalog check. One-time install per clone:

git config --local core.hooksPath .githooks

To uninstall: git config --local --unset core.hooksPath.

Why a wrapper? When core.hooksPath is set globally (e.g. by Alibaba's AK scanner at ~/.aliyunAKScanHook/hooks), the per-repo .git/hooks/ directory is ignored. We have to set core.hooksPath to a checked-in directory (.githooks) and explicitly re-invoke the global hook from inside the wrapper so the AK scan still runs.

Project Structure

agent-skills/
├── skills/                          # Source skills
│   ├── hologres-cli/
│   ├── hologres-query-optimizer/
│   ├── hologres-slow-query-analysis/
│   ├── hologres-schema-generator/
│   ├── hologres-privileges/
│   ├── hologres-uv-compute/
│   ├── hologres-bsi-profile-analysis/
│   ├── hologres-ad-campaign/
│   ├── hologres-instance-health-analyse/
│   ├── hologres-diagnosis-cpu/
│   ├── hologres-diagnosis-memory/
│   ├── hologres-daily-report/
│   └── hologres-knowledge-base/
├── src/
│   └── holo_plugin_installer/
│       ├── __init__.py
│       └── main.py
├── pyproject.toml
├── MANIFEST.in
├── upload_to_pypi.py
├── publish_to_aone.py
├── README.md
└── README_CN.md

License

Apache License 2.0 — Copyright 2026 Alibaba Cloud

Metadata

Release files for hologres-agent-skills 0.2.5

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for hologres-agent-skills 0.2.5
File Size Uploaded Provenance Actions
hologres_agent_skills-0.2.5.tar.gz 379.1 kB Without provenance Details for hologres_agent_skills-0.2.5.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for hologres-agent-skills 0.2.5
File Interpreter ABI Platform Actions
hologres_agent_skills-0.2.5-py3-none-any.whl Python 3 none any Details for hologres_agent_skills-0.2.5-py3-none-any.whl

Total release size: 631.6 kB

Release files / hologres_agent_skills-0.2.5.tar.gz

Download URL hologres_agent_skills-0.2.5.tar.gz
Size 379.1 kB
Tags Source
SHA-256 checksum
How to use checksums
7404475bca9c9a6feea3c271121f279ef6a0418f43824ce4610b514baf6055f8
BLAKE2b-256 checksum
How to use checksums
ceb3f20866ff9696618793b0afc17d81ba7dade7369164a1a29bd321106d4d80
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.11.25 {"installer":{"name":"uv","version":"0.11.25","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

Release files / hologres_agent_skills-0.2.5-py3-none-any.whl

Download URL hologres_agent_skills-0.2.5-py3-none-any.whl
Size 252.5 kB
Tags Python 3
SHA-256 checksum
How to use checksums
f4ea88a99b1d5e0779acbdabddc0b1024a4da6f924799e33ab14b89ad8f13bef
BLAKE2b-256 checksum
How to use checksums
d07e1e145ea8d60ebcf70b9f72951fd7482c91544a41c96aa597631386c382d9
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.11.25 {"installer":{"name":"uv","version":"0.11.25","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

Security for hologres-agent-skills 0.2.5

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page