Persistent memory and context management for AI agents using OpenContext. Keep context across sessions/repos/dates, store conclusions, and provide document search workflows.
Install with Tessl CLI
npx tessl i github:supercent-io/skills-template --skill opencontext70
Quality
55%
Does it follow best practices?
Impact
96%
2.13xAverage score across 3 eval scenarios
Optimize this skill with Tessl
npx tessl skill review --optimize ./.agent-skills/opencontext/SKILL.mdGive your AI assistant persistent memory. Stop repeating explanations, and build smarter.
When working with an AI assistant, context gets lost (across sessions, repos, and dates). You end up repeating background, re-explaining decisions, and sometimes the assistant continues with incorrect assumptions.
OpenContext is a lightweight personal context/knowledge store for AI assistants.
[Load context] → [Do work] → [Store conclusions]| Item | Path |
|---|---|
| Contexts | ~/.opencontext/contexts |
| Database | ~/.opencontext/opencontext.db |
npm install -g @aicontextlab/cli
# Or use npx
npx @aicontextlab/cli <command>cd your-project
oc initWhat oc init does:
| Command | Purpose |
|---|---|
/opencontext-help | When you don't know where to start |
/opencontext-context | (Recommended default) Load background before work |
/opencontext-search | Search existing documents |
/opencontext-create | Create a new document/idea |
/opencontext-iterate | Store conclusions and citations |
# Slash Commands
Cursor: ~/.cursor/commands
Claude Code: ~/.claude/commands
# Skills
Cursor: ~/.cursor/skills/opencontext-*/SKILL.md
Claude Code: ~/.claude/skills/opencontext-*/SKILL.md
Codex: ~/.codex/skills/opencontext-*/SKILL.md
# MCP Config
Cursor: ~/.cursor/mcp.json
Claude Code: ~/.claude/mcp.json# List folders
oc folder ls --all
# Create folder
oc folder create project-a -d "My project"
# Create document
oc doc create project-a design.md -d "Design doc"
# List documents
oc doc ls project-a# Search (keyword/hybrid/vector)
oc search "your query" --mode keyword --format json
# Generate a manifest (list of files the AI should read)
oc context manifest project-a --limit 10| Mode | Description | Requirements |
|---|---|---|
--mode keyword | Keyword-based search | No embeddings required |
--mode vector | Vector search | Embeddings + index required |
--mode hybrid | Hybrid (default) | Embeddings + index required |
# Set API key
oc config set EMBEDDING_API_KEY "<<your_key>>"
# (Optional) Set base URL
oc config set EMBEDDING_API_BASE "https://api.openai.com/v1"
# (Optional) Set model
oc config set EMBEDDING_MODEL "text-embedding-3-small"
# Build index
oc index buildoc_list_folders # List folders
oc_list_docs # List documents
oc_manifest # Generate manifest
oc_search # Search documents
oc_create_doc # Create document
oc_get_link # Generate stable link# Gemini: large-scale analysis
ask-gemini "Analyze the structure of the entire codebase"
# Codex: run commands
shell "docker-compose up -d"
# OpenContext: store results
oc doc create project-a conclusions.md -d "Analysis conclusions"[Claude] Plan
↓
[Gemini] Analysis/research + OpenContext search
↓
[Claude] Write code
↓
[Codex] Run/test
↓
[Claude] Synthesize results + store in OpenContext# 1. [Claude] Design API spec using the skill
/opencontext-context # Load project background
# 2. [Gemini] Analyze a large codebase
ask-gemini "@src/ Analyze existing API patterns"
# 3. [Claude] Implement code based on the analysis
# (Use context loaded from OpenContext)
# 4. [Codex] Test and build
shell "npm test && npm run build"
# 5. [Claude] Create final report + store conclusions
/opencontext-iterate # Store decisions and lessons learned/opencontext-context/opencontext-search/opencontext-iterateKeep links stable across renames/moves by referencing document IDs:
[label](oc://doc/<stable_id>)oc doc link <doc_path>oc_get_link doc_path="Product/api-spec"Citation features:
| Action | How | Result |
|---|---|---|
| Cite text snippet | Select text → right-click → "Copy Citation" | Agent reads the snippet + source |
| Cite document | Click the citation icon next to the document title | Agent reads the full document + obtains stable_id |
| Cite folder | Right-click folder → "Copy Folder Citation" | Agent bulk-reads all docs in the folder |
oc ui
# Default URL: http://127.0.0.1:4321Before: /opencontext-context (load background)
During: /opencontext-search (search)
After: /opencontext-iterate (store)oc init # Initialize project
oc folder ls --all # List folders
oc doc ls <folder> # List documents
oc search "query" # Search
oc doc create ... # Create documentoc_list_folders list folders
oc_list_docs list documents
oc_search search
oc_manifest manifest
oc_create_doc create document
oc_get_link generate link~/.opencontext/contexts context store
~/.opencontext/opencontext.db database2311988
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.