Pi-based agentic document redaction: local Docker orchestration and Hugging Face Space packaging.
Supports three orchestration backends via AGENT_ORCHESTRATOR (pi default, langgraph, agentcore). See pi/agent/README.md.
| Path | Purpose |
|---|---|
shared/ |
Gradio UI, runtime factory, session workspace, remote redaction helpers |
pi/ |
Pi RPC client, Pi config generation, skills sync, agent/*.json templates |
redaction_langgraph/ |
LangGraph ReAct agent (curated tools, no shell) |
agentcore/ |
Bedrock AgentCore runtime + Gradio AgentCore clients + install guide |
eval/ |
Arize AX / Phoenix tracing for LangGraph |
pi-agent/ |
Docker image (dev + runtime targets), sync script, and manifest |
requirements_agent.txt |
Python deps for agent-redact Docker images |
Per-user output isolation uses Gradio session_hash subfolders under AGENT_WORKSPACE_DIR (see agent-redact/shared/session_workspace.py). Enabled by default locally and on HF Spaces. Set AGENT_SESSION_WORKSPACE=false only if you want one shared workspace tree for all sessions.
Local Docker
Use the pi-agent service in docker-compose_llama_agentic.yml (profile 27b_36). See pi/agent/README.md.
Hugging Face Space
Build from repo root:
# Production (HF Space / ECS)
docker build -f agent-redact/pi-agent/Dockerfile --target runtime .
# Local compose (bind-mounted repo)
docker build -f agent-redact/pi-agent/Dockerfile --target dev .
Sync to Space on pushes to dev via .github/workflows/sync-pi-agent-space.yml.