PicoOraClaw: Ultra-Efficient AI Assistant in Go + ollama + Oracle AI Database based on PicoClaw

$10 Hardware · 10MB RAM · 1s Boot · Oracle AI Vector Search · ollama-based

PicoOraClaw is a fork of PicoClaw that adds Oracle AI Database as a backend for persistent storage and semantic vector search, as well as ollama for LLM communication. The agent remembers facts and recalls them by meaning using in-database ONNX embeddings — no external embedding API required.

_{Under 10MB RAM — runs on $10 hardware}

Architecture at a Glance

Full interactive presentation: Open picooraclaw-presentation.html in your browser for all 24 slides with animations and keyboard navigation.

Title	By the Numbers
23 Packages	10 Channels
Hardware Tools	Oracle Schema

Installation

One-command install — clone, configure, and run in a single step:

curl -fsSL https://raw.githubusercontent.com/jasperan/picooraclaw/main/install.sh | bash

Advanced options

Override install location:

PROJECT_DIR=/opt/myapp curl -fsSL https://raw.githubusercontent.com/jasperan/picooraclaw/main/install.sh | bash

Or install manually:

git clone https://github.com/jasperan/picooraclaw.git
cd picooraclaw
# See below for setup instructions

🦾 Demonstration

🛠️ Standard AI Workflows

🧩 Full-Stack Engineering	🧠 Oracle AI Memory	🔎 Web Search & Learning

Develop · Deploy · Scale	Remember · Recall · Persist	Discover · Insights · Trends

⏰ Scheduled Tasks & Reminders

Set reminders, run recurring tasks, automate workflows — scheduled jobs are stored persistently in Oracle AI Database with full ACID guarantees.

Quickstart (5 minutes)

Everything you need: Go 1.24+, Ollama and Docker (for Oracle AI Database 26ai Free).

Step 1: Build

git clone https://github.com/jasperan/picooraclaw.git
cd picooraclaw
make build

Step 2: Initialize

./build/picooraclaw onboard

Step 3: Start Ollama and pull a model

# Install Ollama if needed: https://ollama.com/download
ollama pull qwen3:latest

Step 4: Configure for Ollama

Edit ~/.picooraclaw/config.json:

{
  "agents": {
    "defaults": {
      "provider": "ollama",
      "model": "qwen3:latest",
      "max_tokens": 8192,
      "temperature": 0.7
    }
  },
  "providers": {
    "ollama": {
      "api_key": "",
      "api_base": "http://localhost:11434/v1"
    }
  }
}

Step 5: Chat

# One-shot
./build/picooraclaw agent -m "Hello!"

# Interactive mode
./build/picooraclaw agent

That's it — you have a working AI assistant with local inference. No API keys, no cloud dependency.

Deploy to Oracle Cloud (One-Click)

Deploy a fully configured PicoOraClaw instance on OCI with Oracle AI Database, Ollama, and the gateway — all automated.

What gets deployed:

OCI Compute instance (shape of your choice, ARM A1.Flex is Always Free)
Ollama with gemma3:270m pre-loaded for CPU inference
Oracle AI Database 26ai Free container by default (or optional Autonomous AI Database when toggled)
PicoOraClaw gateway running as a systemd service

After deployment (~5-8 min for setup to complete):

# Check setup progress
ssh opc@<public_ip> -t 'tail -f /var/log/picooraclaw-setup.log'

# Start chatting
ssh opc@<public_ip> -t picooraclaw agent

# Check gateway health
curl http://<public_ip>:18790/health

Add Oracle AI Vector Search

Oracle gives you persistent storage, semantic memory (remember/recall by meaning), and crash-safe ACID transactions. Without it, storage is file-based.

Run the setup script — it handles everything automatically:

./scripts/setup-oracle.sh [optional-password]

This single script:

Pulls and starts the Oracle AI Database 26ai Free container
Waits for the database to be ready
Creates the picooraclaw database user with the required grants
Patches your ~/.picooraclaw/config.json with the Oracle connection settings
Runs picooraclaw setup-oracle to initialize the schema and load the ONNX embedding model

Expected output when complete:

── Step 4/4: Schema + ONNX model ─────────────────────────────────────────
  Running picooraclaw setup-oracle...
✓ Connected to Oracle AI Database
✓ Schema initialized (8 tables with PICO_ prefix)
✓ ONNX model 'ALL_MINILM_L12_V2' already loaded
✓ VECTOR_EMBEDDING() test passed
✓ Prompts seeded from workspace

════════════════════════════════════════════════════════
  Oracle AI Database setup complete!
  Test with:
    ./build/picooraclaw agent -m "Remember that I love Go"
    ./build/picooraclaw agent -m "What language do I like?"
    ./build/picooraclaw oracle-inspect
════════════════════════════════════════════════════════

Step 5: Test semantic memory

# Store a fact
./build/picooraclaw agent -m "Remember that my favorite language is Go"

# Recall by meaning (not keywords)
./build/picooraclaw agent -m "What programming language do I prefer?"

The second command finds the stored memory via cosine similarity on 384-dimensional vectors — no keyword matching.

Step 6: Inspect what's stored

The oracle-inspect command lets you view everything stored in Oracle without writing SQL.

picooraclaw oracle-inspect [table] [options]

Tables: memories, sessions, transcripts, state, notes, prompts, config, meta Options: -n <limit> max rows (default 20), -s <text> semantic search (memories only)

Overview dashboard (no arguments)

./build/picooraclaw oracle-inspect

=============================================================
  PicoOraClaw Oracle AI Database Inspector
=============================================================

  Table                  Rows
  ─────────────────────  ────
  Memories                  20  ████████████████████
  Sessions                   4  ████
  Transcripts                6  ██████
  State                      8  ████████
  Daily Notes                3  ███
  Prompts                    4  ████
  Config                     2  ██
  Meta                       1  █
  ─────────────────────  ────
  Total                     48

  Recent Memories (last 5):
  ─────────────────────────────────────────────────────────
  2026-02-19 04:13  0.6 [preference]  Design docs and plan files go in docs/plans/ ...
  2026-02-19 04:13  0.6 [interest]  User is interested in IoT and embedded systems ...
  2026-02-19 04:13  0.7 [fact]  The Oracle ONNX embedding model used is ALL_MINILM_L12_V2 ...
  2026-02-19 04:13  0.7 [preference]  For complex multi-step tasks, use multi-agent parallel ...
  2026-02-19 04:13  0.8 [preference]  User prefers concise communication ...

  Recent Transcripts (last 5):
  ─────────────────────────────────────────────────────────
  2026-02-19 04:13  assistant   [discord:dev-channel]  I've stored that in memory ...
  2026-02-19 04:13  user        [discord:dev-channel]  Remember that the next release ...
  2026-02-19 04:13  assistant   [telegram:user123]  Oracle Database connection is active ...
  2026-02-19 04:13  user        [telegram:user123]  What's the status of the Oracle connection?
  2026-02-19 04:13  assistant   [cli:repl-session]  I can help with that ...

  Recent Sessions (last 5):
  ─────────────────────────────────────────────────────────
  2026-02-19 04:13  discord:dev-channel             (no summary)
  2026-02-19 04:13  telegram:user123                (no summary)
  2026-02-19 04:13  cli:repl-session                (no summary)
  2026-02-18 06:07  cli:default                     **Cohesive Summary:** The user and
  assistant engaged in a playful exchange starting with "Pong!" ...

  Recent State Entries (last 5):
  ─────────────────────────────────────────────────────────
  2026-02-19 04:13  total_conversations       = 42
  2026-02-19 04:13  agent_mode                = interactive
  2026-02-19 04:13  user_timezone             = America/Los_Angeles
  2026-02-19 04:13  user_name                 = jasperan
  2026-02-19 04:13  last_chat_id              = repl-session

  Recent Daily Notes (last 5):
  ─────────────────────────────────────────────────────────
  2026-02-19  (updated 2026-02-19 04:13)  # 2026-02-19 ...
  2026-02-18  (updated 2026-02-19 04:13)  # 2026-02-18 ...
  2026-02-17  (updated 2026-02-19 04:13)  # 2026-02-17 ...

  System Prompts (last 5):
  ─────────────────────────────────────────────────────────
  2026-02-18 08:38  AGENT                      357 chars
  2026-02-18 08:38  USER                       365 chars
  2026-02-18 08:38  SOUL                       296 chars
  2026-02-18 08:38  IDENTITY                  1271 chars

  Config Entries (last 5):
  ─────────────────────────────────────────────────────────
  2026-02-19 04:13  last_seed                 = 2026-02-19T04:13:34Z
  2026-02-19 04:13  full_config               = { "llm": {...}, "oracle": {...}, ... }

  Schema Metadata:
  ─────────────────────────────────────────────────────────
  2026-02-18 08:38  schema_version                 = 1.0.0

  Tip: Run 'picooraclaw oracle-inspect <table>' for details
       Run 'picooraclaw oracle-inspect memories -s "query"' for semantic search

List all memories

./build/picooraclaw oracle-inspect memories

  All Memories
  ─────────────────────────────────────────────────────────

  ID: faffd019  Vector: yes
  Created: 2026-02-19 04:13  Importance: 0.9  Category: preference  Accessed: 0x
  Content: User prefers Oracle Database as the primary database. They work at Oracle
  and prefer Oracle AI Vector Search for embeddings.

  ID: 0e39036f  Vector: yes
  Created: 2026-02-19 04:13  Importance: 0.8  Category: preference  Accessed: 0x
  Content: Go is the user's primary programming language. They use Go 1.24 and target
  embedded Linux devices (RISC-V, ARM64, x86_64).

  ID: 7aca4a7b  Vector: yes
  Created: 2026-02-19 04:13  Importance: 0.8  Category: preference  Accessed: 0x
  Content: User prefers Ollama as the open-source LLM framework for local inference.

  ID: 27c2473c  Vector: yes
  Created: 2026-02-19 04:13  Importance: 0.7  Category: interest  Accessed: 0x
  Content: User is interested in RAG (Retrieval-Augmented Generation) using LlamaIndex
  and LangChain frameworks.

  ...

Semantic search over memories

./build/picooraclaw oracle-inspect memories -s "what does the user like to program in"

  Semantic Search: "what does the user like to program in"
  ─────────────────────────────────────────────────────────

  [ 61.3% match]  ID: 383ff5d3
  Created: 2026-02-16 06:13  Importance: 0.7  Category: preference  Accessed: 0x
  Content: I prefer Python and Go for programming

  [ 60.7% match]  ID: 0e74a94c
  Created: 2026-02-18 02:20  Importance: 0.7  Category: preference  Accessed: 0x
  Content: my favorite programming language is Go

  [ 40.1% match]  ID: 0e39036f
  Created: 2026-02-19 04:13  Importance: 0.8  Category: preference  Accessed: 0x
  Content: Go is the user's primary programming language. They use Go 1.24 and target
  embedded Linux devices (RISC-V, ARM64, x86_64).

  [ 30.9% match]  ID: 22b84dba
  Created: 2026-02-16 06:12  Importance: 0.7  Category: employment  Accessed: 1x
  Content: I work at Oracle as a developer

  ...

Inspect sessions

./build/picooraclaw oracle-inspect sessions

  Chat Sessions
  ─────────────────────────────────────────────────────────

  Session: discord:dev-channel
  Created: 2026-02-19 04:13  Updated: 2026-02-19 04:13  Messages size: 673 bytes

  Session: telegram:user123
  Created: 2026-02-19 04:13  Updated: 2026-02-19 04:13  Messages size: 304 bytes

  Session: cli:repl-session
  Created: 2026-02-19 04:13  Updated: 2026-02-19 04:13  Messages size: 982 bytes

  Session: cli:default
  Created: 2026-02-16 06:12  Updated: 2026-02-18 06:07  Messages size: 2848 bytes
  Summary: **Cohesive Summary:** The user and assistant engaged in a playful
  exchange starting with "Pong!" The assistant recalled the user's role as an
  Oracle developer and their preference for Go ...

Inspect agent state

./build/picooraclaw oracle-inspect state

  Agent State (Key-Value)
  ─────────────────────────────────────────────────────────
  agent_mode                     = interactive                    (2026-02-19 04:13)
  last_channel                   = cli                            (2026-02-19 04:13)
  last_chat_id                   = repl-session                   (2026-02-19 04:13)
  last_model                     = gpt-4o-mini                    (2026-02-19 04:13)
  tools_used_count               = 187                            (2026-02-19 04:13)
  total_conversations            = 42                             (2026-02-19 04:13)
  user_name                      = jasperan                       (2026-02-19 04:13)
  user_timezone                  = America/Los_Angeles            (2026-02-19 04:13)

Inspect daily notes

./build/picooraclaw oracle-inspect notes

  Daily Notes
  ─────────────────────────────────────────────────────────

  Date: 2026-02-19  ID: 2ccb7e70  Vector: yes  Updated: 2026-02-19 04:13
  Content: # 2026-02-19
  ## Development Progress
  - Implemented seed-demo command for Oracle data population
  - Tested vector embeddings with ALL_MINILM_L12_V2 model
  - Fixed session serialization edge case with empty tool calls

  Date: 2026-02-18  ID: 29a60563  Vector: yes  Updated: 2026-02-19 04:13
  Content: # 2026-02-18
  ## Testing & Debugging
  - Ran full test suite: all 47 tests passing
  - Profiled memory usage on RISC-V board: 8.2MB peak
  - Verified Oracle connection pooling under concurrent load

  Date: 2026-02-17  ID: f4cfb628  Vector: no  Updated: 2026-02-19 04:13
  Content: # 2026-02-17
  ## Architecture Planning
  - Designed transcript storage schema for PICO_TRANSCRIPTS table
  - Sketched WhatsApp channel adapter following existing Telegram pattern
  - Reviewed LangChain-OracleDB integration for RAG pipeline

Inspect transcripts

./build/picooraclaw oracle-inspect transcripts

  Conversation Transcripts
  ─────────────────────────────────────────────────────────
  2026-02-19 04:13  #2  assistant  [discord:dev-channel]   I've stored that in memory ...
  2026-02-19 04:13  #1  user       [discord:dev-channel]   Remember that the next release ...
  2026-02-19 04:13  #2  assistant  [telegram:user123]      Oracle Database connection is active ...
  2026-02-19 04:13  #1  user       [telegram:user123]      What's the status of the Oracle connection?
  2026-02-19 04:13  #2  assistant  [cli:repl-session]      I can help with that ...
  2026-02-19 04:13  #1  user       [cli:repl-session]      Can you help me add a new tool ...

Inspect stored config

./build/picooraclaw oracle-inspect config

  Stored Config
  ─────────────────────────────────────────────────────────
  full_config                    = {
  "llm": {"provider": "openai-compatible", "model": "gpt-4o-mini", ...},
  "oracle": {"enabled": true, "onnx_model": "ALL_MINILM_L12_V2"},
  "channels": {"telegram": {"enabled": true}, "discord": {"enabled": true}},
  "agent": {"max_tool_iterations": 10, "context_window": 8192}
}  (2026-02-19 04:13)
  last_seed                      = 2026-02-19T04:13:34Z       (2026-02-19 04:13)

View a system prompt in full

./build/picooraclaw oracle-inspect prompts IDENTITY
./build/picooraclaw oracle-inspect prompts SOUL

Schema metadata, ONNX models, and vector indexes

./build/picooraclaw oracle-inspect meta

  Schema Metadata
  ─────────────────────────────────────────────────────────
  schema_version                 = 1.0.0

  ONNX Models
  ─────────────────────────────────────────────────────────
  ALL_MINILM_L12_V2          EMBEDDING        ONNX

  Vector Indexes
  ─────────────────────────────────────────────────────────
  IDX_PICO_DAILY_NOTES_VEC        on PICO_DAILY_NOTES
  IDX_PICO_MEMORIES_VEC           on PICO_MEMORIES

CLI Reference

Command	Description
`picooraclaw onboard`	Initialize config and workspace
`picooraclaw agent -m "..."`	One-shot chat
`picooraclaw agent`	Interactive chat mode
`picooraclaw gateway`	Start long-running service with channels
`picooraclaw status`	Show status
`picooraclaw setup-oracle`	Initialize Oracle schema + ONNX model
`picooraclaw oracle-inspect`	Inspect data stored in Oracle
`picooraclaw oracle-inspect memories -s "query"`	Semantic search over memories
`picooraclaw seed-demo`	Populate Oracle with realistic demo data
`picooraclaw cron list`	List scheduled jobs
`picooraclaw skills list`	List installed skills

How Oracle Storage Works

                           ┌──────────────────────────────────────────┐
                           │         Oracle AI Database               │
                           │                                          │
  picooraclaw binary       │  ┌──────────────┐  ┌──────────────────┐ │
  ┌───────────────────┐    │  │ PICO_MEMORIES │  │ PICO_DAILY_NOTES │ │
  │  AgentLoop        │    │  │  + VECTOR idx │  │  + VECTOR idx    │ │
  │  ├─ SessionStore ──────│──│──────────────┐│  └──────────────────┘ │
  │  ├─ StateStore   ──────│──│ PICO_SESSIONS││                       │
  │  ├─ MemoryStore  ──────│──│ PICO_STATE   ││  ┌──────────────────┐ │
  │  ├─ PromptStore  ──────│──│ PICO_PROMPTS ││  │ ALL_MINILM_L12_V2│ │
  │  ├─ ConfigStore  ──────│──│ PICO_CONFIG  ││  │   (ONNX model)   │ │
  │  └─ Tools:       │    │  │ PICO_META    ││  │  384-dim vectors  │ │
  │     ├─ remember  ──────│──│ PICO_TRANS.  ││  └──────────────────┘ │
  │     └─ recall    ──────│──└──────────────┘│                       │
  └───────────────────┘    │   go-ora v2.9.0  │                       │
         (pure Go)         │   (pure Go driver)│                       │
                           └──────────────────────────────────────────┘

Table	Purpose
`PICO_MEMORIES`	Long-term memory with 384-dim vector embeddings for semantic search
`PICO_SESSIONS`	Chat history per channel
`PICO_TRANSCRIPTS`	Full conversation audit log
`PICO_STATE`	Agent key-value state
`PICO_DAILY_NOTES`	Daily journal entries with vector embeddings
`PICO_PROMPTS`	System prompts (IDENTITY.md, SOUL.md, etc.)
`PICO_CONFIG`	Runtime configuration
`PICO_META`	Schema versioning metadata

The remember tool stores text + vector embedding via VECTOR_EMBEDDING(ALL_MINILM_L12_V2 USING :text AS DATA). The recall tool searches by cosine similarity via VECTOR_DISTANCE(). Results with < 30% similarity are filtered out.

Using a Cloud LLM Instead of Ollama

If you prefer a cloud provider, set provider and add your API key:

OpenRouter (access to all models)

{
  "agents": {
    "defaults": {
      "provider": "openrouter",
      "model": "anthropic/claude-opus-4-5"
    }
  },
  "providers": {
    "openrouter": {
      "api_key": "sk-or-v1-xxx",  # pragma: allowlist secret
      "api_base": "https://openrouter.ai/api/v1"
    }
  }
}

Get a key at openrouter.ai/keys (200K free tokens/month).

Zhipu (best for Chinese users)

{
  "agents": {
    "defaults": {
      "provider": "zhipu",
      "model": "glm-4.7"
    }
  },
  "providers": {
    "zhipu": {
      "api_key": "your-key",  # pragma: allowlist secret
      "api_base": "https://open.bigmodel.cn/api/paas/v4"
    }
  }
}

Get a key at bigmodel.cn.

All supported providers

Provider	Purpose	Get API Key
`ollama`	Local inference (recommended)	ollama.com
`openrouter`	Access to all models	openrouter.ai
`zhipu`	Zhipu/GLM models	bigmodel.cn
`anthropic`	Claude models	console.anthropic.com
`openai`	GPT models	platform.openai.com
`gemini`	Gemini models	aistudio.google.com
`deepseek`	DeepSeek models	platform.deepseek.com
`groq`	Fast inference + voice transcription	console.groq.com

OCI Generative AI (Optional)

PicoOraClaw can optionally use OCI Generative AI as an LLM backend via the oci-openai Python library. This is not required — Ollama remains the default and recommended LLM backend.

Why OCI GenAI?

Enterprise models — Access xAI Grok, Meta Llama, Cohere, and other models through OCI
OCI-native auth — Uses your existing ~/.oci/config profile (no separate API keys)
Same region as your database — Run inference and storage in the same OCI region

Setup

Install the OCI GenAI proxy:

cd oci-genai
pip install -r requirements.txt

Configure OCI credentials (~/.oci/config):

[DEFAULT]
user=ocid1.user.oc1..aaaaaaaaexample
fingerprint=aa:bb:cc:dd:ee:ff:00:11:22:33:44:55:66:77:88:99
tenancy=ocid1.tenancy.oc1..aaaaaaaaexample
region=us-chicago-1
key_file=~/.oci/oci_api_key.pem

Set environment variables:

export OCI_PROFILE=DEFAULT
export OCI_REGION=us-chicago-1
export OCI_COMPARTMENT_ID=ocid1.compartment.oc1..your-compartment-ocid

Start the OCI GenAI proxy:

cd oci-genai
python proxy.py
# Proxy runs at http://localhost:9999/v1

Configure PicoOraClaw (~/.picooraclaw/config.json):

{
  "provider": "openai",
  "api_base": "http://localhost:9999/v1",
  "api_key": "oci-genai",  # pragma: allowlist secret
  "model": "meta.llama-3.3-70b-instruct"
}

See oci-genai/README.md for full documentation.

Chat Channels

Connect PicoOraClaw to Telegram, Discord, Slack, DingTalk, LINE, QQ, or Feishu via the gateway command.

Telegram (Recommended)

Message @BotFather on Telegram, send /newbot, copy the token
Add to ~/.picooraclaw/config.json:

{
  "channels": {
    "telegram": {
      "enabled": true,
      "token": "YOUR_BOT_TOKEN",
      "allowFrom": ["YOUR_USER_ID"]
    }
  }
}

Run picooraclaw gateway

Discord

Create a bot at discord.com/developers, enable MESSAGE CONTENT INTENT
Add to config:

{
  "channels": {
    "discord": {
      "enabled": true,
      "token": "YOUR_BOT_TOKEN",
      "allowFrom": ["YOUR_USER_ID"]
    }
  }
}

Invite bot with Send Messages + Read Message History permissions
Run picooraclaw gateway

QQ, DingTalk, LINE, Feishu, Slack

See config/config.example.json for the full channel configuration reference. Each channel follows the same pattern:

{
  "channels": {
    "<channel_name>": {
      "enabled": true,
      "<credentials>": "...",
      "allow_from": []
    }
  }
}

Run picooraclaw gateway after configuring.

Oracle on Autonomous AI Database (Cloud, Optional)

ADB wallet-less TLS

{
  "oracle": {
    "enabled": true,
    "mode": "adb",
    "dsn": "(description=(retry_count=20)(retry_delay=3)(address=(protocol=tcps)(port=1522)(host=adb.us-ashburn-1.oraclecloud.com))(connect_data=(service_name=xxx_myatp_low.adb.oraclecloud.com))(security=(ssl_server_dn_match=yes)))",
    "user": "picooraclaw",
    "password": "YourPass123"  # pragma: allowlist secret
  }
}

ADB with mTLS wallet

{
  "oracle": {
    "enabled": true,
    "mode": "adb",
    "host": "adb.us-ashburn-1.oraclecloud.com",
    "port": 1522,
    "service": "xxx_myatp_low.adb.oraclecloud.com",
    "walletPath": "/path/to/wallet",
    "user": "picooraclaw",
    "password": "YourPass123"  # pragma: allowlist secret
  }
}

Download wallet from OCI Console > Autonomous Database > DB Connection > Download Wallet.

Oracle config reference

Field	Env Variable	Default	Description
`enabled`	`PICO_ORACLE_ENABLED`	`false`	Enable Oracle backend
`mode`	`PICO_ORACLE_MODE`	`freepdb`	`freepdb` or `adb`
`host`	`PICO_ORACLE_HOST`	`localhost`	Oracle host
`port`	`PICO_ORACLE_PORT`	`1521`	Listener port
`service`	`PICO_ORACLE_SERVICE`	`FREEPDB1`	Service name
`user`	`PICO_ORACLE_USER`	`picooraclaw`	DB username
`password`	`PICO_ORACLE_PASSWORD`	—	DB password
`dsn`	`PICO_ORACLE_DSN`	—	Full DSN (ADB wallet-less)
`walletPath`	`PICO_ORACLE_WALLET_PATH`	—	Wallet directory (ADB mTLS)
`onnxModel`	`PICO_ORACLE_ONNX_MODEL`	`ALL_MINILM_L12_V2`	ONNX model for embeddings
`agentId`	`PICO_ORACLE_AGENT_ID`	`default`	Multi-agent isolation key

Troubleshooting

Oracle: Connection refused / ORA-12541

docker ps | grep oracle          # Is it running?
docker logs oracle-free          # Wait for "DATABASE IS READY"
ss -tlnp | grep 1521            # Is port 1521 listening?

Oracle: ORA-01017 invalid username/password

docker exec -it oracle-free sqlplus sys/YourPass123@localhost:1521/FREEPDB1 as sysdba
SQL> ALTER USER picooraclaw IDENTIFIED BY NewPassword123;

Oracle: VECTOR_EMBEDDING() returns ORA-04063

ONNX model not loaded. Run picooraclaw setup-oracle or manually:

BEGIN
  DBMS_VECTOR.LOAD_ONNX_MODEL('PICO_ONNX_DIR', 'all_MiniLM_L12_v2.onnx', 'ALL_MINILM_L12_V2');
END;
/

Requires GRANT CREATE MINING MODEL TO picooraclaw; as SYSDBA.

Agent falls back to file-based mode

Oracle is enabled but connection failed at startup. Check:

Is the Oracle container healthy? (docker ps)
Password match between config and ORACLE_PWD?
Service name should be FREEPDB1 (not FREE or XE)

Build Targets

make build          # Build for current platform
make build-all      # Cross-compile: linux/{amd64,arm64,riscv64}, darwin/arm64, windows/amd64
make install        # Build + install to ~/.local/bin
make test           # go test ./...
make fmt            # go fmt ./...
make vet            # go vet ./...

Docker Compose

# Full stack with Oracle
PICO_ORACLE_PASSWORD=YourPass123 docker compose --profile oracle --profile gateway up -d  # pragma: allowlist secret

# Without Oracle
docker compose --profile gateway up -d

# One-shot agent
docker compose run --rm picoclaw-agent -m "What is 2+2?"

Features

Single static binary (~10MB RAM), runs on RISC-V/ARM64/x86_64
Ollama, OpenRouter, Anthropic, OpenAI, Gemini, Zhipu, DeepSeek, Groq providers
Default: Oracle AI Database 26ai Free with AI Vector Search (384-dim ONNX embeddings)
Chat channels: Telegram, Discord, Slack, QQ, DingTalk, LINE, Feishu, WhatsApp
Scheduled tasks via cron expressions
Heartbeat periodic tasks
Skills system (workspace, global, GitHub-hosted)
Security sandbox with workspace restriction
Optional: Oracle Autonomous AI Database for managed cloud deployment
Graceful fallback to file-based storage when Oracle is unavailable

Name		Name	Last commit message	Last commit date
Latest commit History 47 Commits
.serena		.serena
assets		assets
cmd/picooraclaw		cmd/picooraclaw
config		config
deploy/oci		deploy/oci
docs/slides		docs/slides
oci-genai		oci-genai
pkg		pkg
scripts		scripts
workspace		workspace
.dockerignore		.dockerignore
.gitignore		.gitignore
.goreleaser.yaml		.goreleaser.yaml
Dockerfile		Dockerfile
Dockerfile.goreleaser		Dockerfile.goreleaser
LICENSE		LICENSE
Makefile		Makefile
README.ja.md		README.ja.md
README.md		README.md
README.zh.md		README.zh.md
docker-compose.yml		docker-compose.yml
go.mod		go.mod
go.sum		go.sum
install.sh		install.sh
picooraclaw-presentation.html		picooraclaw-presentation.html

Folders and files

Latest commit

History

Repository files navigation