ultron-1.0.0
starsforks
How to install ultron-1.0.0?
modelscope/ultron/ultron-1.0.0
$npx -y skills add modelscope/ultron --skill ultron-1.0.0Ultron: Collective Memory Synchronization System for General-Purpose Agents
SKILL.md
| 1 | # Ultron - Shared Collective Memory & Skills |
| 2 | |
| 3 | **Purpose**: Mount this skill on an AI assistant / agent to gain **multi-agent shared remote memory + skills** capabilities. |
| 4 | |
| 5 | Three core functions: |
| 6 | 1. **Shared Collective Memory**: Upload and retrieve collective experience with automatic deduplication and merging; search returns L0/L1 only, use `get_memory_details` for full text |
| 7 | 2. **Skill Hub**: Semantic search (with intent analysis), distillation from high-frequency memories |
| 8 | 3. **Smart Ingestion**: Pass file paths or text → LLM extracts memories + optionally generates skills |
| 9 | |
| 10 | ## On Every User Request: Retrieve Remote Memory & Skills First |
| 11 | |
| 12 | Whenever the user makes a new request that requires thinking or action, **first** retrieve collective memory and skills from Ultron, then proceed. Pure small talk can be skipped. |
| 13 | |
| 14 | Invocation priority: |
| 15 | |
| 16 | **Search memory first** (most scenarios): Debugging, writing/modifying code, checking configs, life questions, experience-based questions |
| 17 | 1. **`search_memory`**: `detail_level=l0` for quick scanning, or `l1` for an overview |
| 18 | 2. When relevant memories are found in results, use **`get_memory_details`** to fetch the full text before acting. L0/L1 are summaries and typically insufficient to directly guide execution. |
| 19 | |
| 20 | **Search skills when needed**: When a specific tool/workflow/methodology is required, the task involves an unfamiliar domain, or a ready-made solution is desired |
| 21 | 1. **`search_skills`**: `query` = user intent in 10–30 words, `limit` 3–5 |
| 22 | 2. To install a skill → **`install_skill`** (pass `full_name` and `target_dir`) |
| 23 | |
| 24 | For complex tasks, search both memory and skills simultaneously. |
| 25 | |
| 26 | ## Client Invocation |
| 27 | |
| 28 | ```bash |
| 29 | python3 skills/ultron-1.0.0/scripts/ultron_client.py '{"action":"<action>", ...}' |
| 30 | ``` |
| 31 | |
| 32 | ### Available Actions |
| 33 | |
| 34 | | Action | Required Params | Description | |
| 35 | |--------|----------------|-------------| |
| 36 | | `search_skills` | `query` | Semantic skill search, optional `limit` | |
| 37 | | `install_skill` | `full_name`, `target_dir` | Install a skill locally | |
| 38 | | `upload_skill` | `paths` | Upload skills (list of skill directory paths) | |
| 39 | | `upload_memory` | `content` | Upload a memory, optional `context`, `resolution`, `tags` | |
| 40 | | `search_memory` | `query` | Search memories, optional `detail_level`(l0/l1), `limit` | |
| 41 | | `get_memory_details` | `memory_ids` | Fetch full text (full) by ID | |
| 42 | | `ingest` | `paths`, `agent_id` | Pass file/directory paths, LLM auto-extracts memories | |
| 43 | | `ingest_text` | `text` | Pass raw text, LLM extracts memories | |
| 44 | |
| 45 | ## L0/L1/Full Layering |
| 46 | |
| 47 | | Level | Content | Use Case | |
| 48 | |-------|---------|----------| |
| 49 | | **L0** | One-line summary (≤64 tokens) | Quick scanning, list browsing | |
| 50 | | **L1** | Core overview (≤256 tokens) | Initial filtering, context supplementation | |
| 51 | | **Full** | Complete original content | Only via `get_memory_details` | |
| 52 | |
| 53 | ## Core Principle |
| 54 | |
| 55 | **Subjective stays local, objective goes remote**: Personal preferences are stored locally only; objective knowledge such as error resolutions, technical patterns, and life tips are uploaded to Ultron for sharing. Individual medical information and identifiable privacy are never uploaded remotely (see `boundaries.md`). |
| 56 | |
| 57 | ## File Index |
| 58 | |
| 59 | | File | Description | |
| 60 | |------|-------------| |
| 61 | | `setup.md` | Setup guide | |
| 62 | | `operations.md` | Memory operations & upload templates | |
| 63 | | `boundaries.md` | Safety boundaries | |
| 64 | | `scripts/ultron_client.py` | API client | |
| 65 | |
| 66 | ## Environment Variables |
| 67 | |
| 68 | | Variable | Required | Default | Description | |
| 69 | |----------|----------|---------|-------------| |
| 70 | | ULTRON_API_URL | yes | https://writtingforfun-ultron.ms.show | Ultron service API endpoint | |
| 71 | | ULTRON_AGENT_ID | yes | — | Unique identifier (UUID) for the current agent, used for ingest progress isolation | |
People who liked this also viewed
6 picksType
- find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skSkillsJul 2026
- grill-meA relentless interview to sharpen a plan or design.SkillsJul 2026
- vercel-react-best-practicesVercel's official collection of agent skillsSkillsJul 2026
- improve-codebase-architectureScan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.SkillsJul 2026
- setup-matt-pocock-skillsConfigure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.SkillsJul 2026
- triageMove issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.SkillsJul 2026
