Skip to content
Back to skills

Upsonic

ASecurity

"Guides Upsonic Python agent-framework workflows, including agents,

  • 247 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 9, 2026
ai-agentspythongobashfastapirailsapidatabasebackend

Works with

  • cli
  • api
  • mcp

Security analysis

A96/100
  • mediumInstalls packages at runtime which could introduce malicious dependencies

Pro scans all 20 files and shows the line behind each finding

Scanned September 9, 2026

npx -y skills add VectorSpaceLab/AREX-Skill --skill upsonic --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Upsonic?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Upsonic
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/vectorspacelab-upsonic/badge)](https://www.skillsdirectory.com/skills/vectorspacelab-upsonic)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: upsonic
description: "Guides Upsonic Python agent-framework workflows, including agents,
  models, tools, memory, RAG, teams, autonomous agents, CLI projects,
  interfaces, skills, and safety governance."
disable-model-invocation: true
metadata:
  disco-role: operating
license: MIT
---

# Upsonic Repo Skill

Use this skill for tasks that involve the Upsonic Python package or repository. It routes by workflow, not by source-folder shape, so future agents can answer common user requests without reopening the repo.

## Route map

| If the user asks about... | Read |
| --- | --- |
| `Agent`, `Task`, `Direct`, `Graph`, execution, streaming, structured output, run control | [sub-skills/agent-runtime/SKILL.md](sub-skills/agent-runtime/SKILL.md) |
| provider/model strings, provider inference, profiles, model selection, credentials | [sub-skills/models-and-providers/SKILL.md](sub-skills/models-and-providers/SKILL.md) |
| custom tools, `@tool`, `ToolConfig`, MCP, tool orchestration, agent-as-tool | [sub-skills/tools-and-mcp/SKILL.md](sub-skills/tools-and-mcp/SKILL.md) |
| `Chat`, `Memory`, session history, storage backends, user/session persistence | [sub-skills/chat-memory-storage/SKILL.md](sub-skills/chat-memory-storage/SKILL.md) |
| `KnowledgeBase`, loaders, splitters, embeddings, vector DBs, OCR, RAG | [sub-skills/knowledge-rag/SKILL.md](sub-skills/knowledge-rag/SKILL.md) |
| `Team`, `AutonomousAgent`, prebuilt agents, `RalphLoop`, `Simulation` | [sub-skills/teams-autonomous-prebuilt/SKILL.md](sub-skills/teams-autonomous-prebuilt/SKILL.md) |
| safety policies, anonymization, reflection, reliability, eval, tracing | [sub-skills/quality-safety-governance/SKILL.md](sub-skills/quality-safety-governance/SKILL.md) |
| `Skill` / `Skills`, loaders, validation, dependency resolution, caching | [sub-skills/skills-system/SKILL.md](sub-skills/skills-system/SKILL.md) |
| `upsonic` CLI projects, `upsonic_configs.json`, FastAPI `/call`, interfaces | [sub-skills/project-cli-interfaces/SKILL.md](sub-skills/project-cli-interfaces/SKILL.md) |

## Install and smoke check

```bash
python -m pip install upsonic
python scripts/check_upsonic_install.py
```

If you need a backend-specific workflow, install the smallest matching extra instead of all optional groups.

## Shared references

- [references/repo-provenance.md](references/repo-provenance.md) for source snapshot freshness.
- [references/package-overview.md](references/package-overview.md)
- [references/optional-extras.md](references/optional-extras.md)
- [references/optional-extras-snapshot.json](references/optional-extras-snapshot.json)
- [references/troubleshooting.md](references/troubleshooting.md)
- [references/repo-routing-metadata.json](references/repo-routing-metadata.json)

## Shared scripts

- [scripts/check_upsonic_install.py](scripts/check_upsonic_install.py)
- [scripts/list_optional_extras.py](scripts/list_optional_extras.py)

## Guardrails

- Do not assume live API keys, databases, vector stores, OCR engines, or MCP servers are available.
- Do not leak local checkout paths, conda prefixes, or private environment details in runtime guidance.
- Prefer import/signature/CLI checks before optional service-backed checks.
- Keep cross-cutting troubleshooting in the root references and workflow-specific troubleshooting near the owning sub-skill.

Files in this skill

  • SKILL.md3.3 KB
  • references/optional-extras-snapshot.json5.3 KB
  • references/optional-extras.md1.9 KB
  • references/package-overview.md3.9 KB
  • references/repo-provenance.md1.1 KB
  • references/repo-routing-metadata.json480 B
  • references/troubleshooting.md2.2 KB
  • scripts/check_upsonic_install.py3.2 KB
  • scripts/list_optional_extras.py1.9 KB
  • sub-skills/agent-runtime/SKILL.md1.6 KB
  • sub-skills/agent-runtime/references/api-reference.md2.6 KB
  • sub-skills/agent-runtime/references/troubleshooting.md1.5 KB
  • sub-skills/agent-runtime/scripts/inspect_runtime_api.py755 B
  • sub-skills/chat-memory-storage/SKILL.md1.2 KB
  • sub-skills/chat-memory-storage/references/chat-and-memory.md1.4 KB
  • sub-skills/chat-memory-storage/references/storage-backends.md1.1 KB
  • sub-skills/chat-memory-storage/references/troubleshooting.md993 B
  • sub-skills/chat-memory-storage/scripts/check_storage_backends.py1.3 KB
  • sub-skills/knowledge-rag/SKILL.md1.1 KB
  • sub-skills/knowledge-rag/references/backends-and-extras.md874 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.