O'Process

AI-native process classification MCP Server. Query 2,325 processes and 3,910 KPIs from APQC PCF 7.4 + ITIL 4 + SCOR 12.0 + AI-era extensions.

Version: 0.3.0 | MCP Spec: 2025-11-25 | MUST: 11/11 | SHOULD: 12/12 | Coverage: 94.72%

What It Does

O'Process gives AI assistants (Claude, GPT, etc.) real-time access to enterprise process knowledge. Connect it as an MCP Server, then ask natural language questions — the AI will call the right tools automatically.

Core capabilities:

• Process Search — "采购流程有哪些?" → returns matching process nodes with hierarchy, description, and confidence score
• Process Tree Navigation — browse the 5-level taxonomy (L1 categories → L5 activities)
• KPI Recommendations — get metrics for any process node (name, unit, formula, direction)
• Role-Process Mapping — "HRBP manages which processes?" → ranked list with confidence scores
• Process Comparison — side-by-side diff of 2+ process nodes across all attributes
• Responsibility Document — generate complete job descriptions with provenance appendix

Why It Matters

Without O'Process	With O'Process
Manually search APQC PCF Excel (1921 rows)	Natural language query, instant results
Guess which KPIs apply to a process	Structured KPI suggestions from 3910 metrics
Write job descriptions from scratch	Auto-generated with process-backed provenance
Cross-reference APQC + ITIL + SCOR manually	Unified 2325-node taxonomy, one query

Use Cases

Management Consulting — Process diagnostics. A manufacturing company's delivery cycle is 30% slower than competitors. Use search_process to locate SCOR Plan/Deliver/Make nodes, then get_kpi_suggestions to build a measurement framework.

HR Digital Transformation — Role-process mapping. CHRO needs to know what processes HR actually owns. Use get_process_tree on node 7.0 (Human Capital) to get the full L1→L4 hierarchy, then map_role_to_processes to map "HRBP" to standard processes.

Legal Due Diligence — Compliance audit. Cross-border M&A requires checking 12+ regulatory domains. Use search_process to locate relevant PCF nodes (corporate governance, tax, labor, environmental), then compare_processes to identify coverage gaps.

Internal Audit — KPI system design. Use get_kpi_suggestions for each process node, review coverage across efficiency/quality/cost/timeliness dimensions, identify missing metrics.

Quick Start

# Install
uv sync

# Run MCP Server (stdio — default)
uv run python -m oprocess.server

# Run with SSE transport
uv run python -m oprocess.server --transport sse --port 8000

# Run with streamable-http transport
uv run python -m oprocess.server --transport streamable-http --port 8000

Claude Desktop Configuration

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "oprocess": {
      "command": "uv",
      "args": ["run", "python", "-m", "oprocess.server"],
      "cwd": "/path/to/O-Process"
    }
  }
}

Tools

8 MCP tools with full input validation (Pydantic Annotated[..., Field(...)]):

Tool	Description	Key Parameters
search_process	Semantic search for process nodes	query (1-500 chars), lang (zh/en), limit (1-50), level (1-5)
get_process_tree	Get process subtree with children	process_id (e.g. "1.0"), max_depth (1-5)
get_kpi_suggestions	Get KPIs for a process node	process_id
compare_processes	Compare 2+ process nodes side-by-side	process_ids (comma-separated, 2+)
get_responsibilities	Generate role responsibilities	process_id, lang, output_format (json/markdown)
map_role_to_processes	Map job role to relevant processes	role_description (1-500 chars), lang, limit, industry
export_responsibility_doc	Export full responsibility document	process_ids (1+), lang, role_name
health_check	Health check — server status and data counts	(none)

All tools return ToolResponse JSON with result, provenance_chain, session_id, and response_ms.

Invalid inputs raise ToolError with descriptive messages. All tools are annotated with readOnlyHint and idempotentHint.

Prompts

3 guided prompt templates for common workflows:

Prompt	Description	Parameters
analyze_process	Step-by-step process analysis workflow	process_id, lang
generate_job_description	Role responsibility document generation	process_ids, role_name, lang
kpi_review	KPI review and gap analysis workflow	process_id, lang

Resources

6 MCP resources for direct data access:

URI	Description
oprocess://process/{id}	Complete process node data
oprocess://category/list	All L1 process categories
oprocess://role/{role_name}	Process mappings for a role
oprocess://audit/session/{id}	Audit log for a session
oprocess://schema/sqlite	SQLite schema definition
oprocess://stats	Framework statistics

Authentication

HTTP transports (SSE, streamable-http) support Bearer token authentication:

# Set API key
export OPROCESS_API_KEY="your-secret-key"

# Client requests must include:
#   Authorization: Bearer your-secret-key

When OPROCESS_API_KEY is not set, authentication is disabled (safe for stdio mode).

Environment Variables

Variable	Required	Description
GOOGLE_API_KEY	No	Enables semantic vector search (gemini-embedding-001). Without it, search falls back to SQL LIKE matching — all features still work.
OPROCESS_API_KEY	No	Bearer token for SSE/HTTP transports. Not needed for stdio mode.
OPROCESS_ALLOWED_ORIGINS	No	Comma-separated allowed origins for CORS (e.g. http://localhost:3000,https://app.example.com)
LOG_LEVEL	No	Logging level (default: WARNING)

No API key is required to run the server. All 8 tools work out of the box. Setting GOOGLE_API_KEY upgrades search_process and map_role_to_processes from text matching to semantic vector search.

Logging

Structured logging via Python logging module:

# Set log level (default: WARNING)
export LOG_LEVEL=INFO   # DEBUG, INFO, WARNING, ERROR

# Logs include: tool name, session ID, response time (ms)

Configuration

Server behavior can be tuned via [tool.oprocess] in pyproject.toml:

Key	Default	Description
boundary_threshold	0.45	Cosine distance threshold for BoundaryResponse
audit_log_enabled	true	Enable/disable SessionAuditLog
default_language	"zh"	Default language (zh/en)
rate_limit_max_calls	60	Max tool calls per client per window
rate_limit_window_seconds	60	Rate limit window duration (seconds)

Rate limiting is enforced per-client via RateLimitMiddleware. Exceeding the limit returns HTTP 429.

Governance-Lite

Transparent governance layer (non-blocking):

• SessionAuditLog -- Append-only invocation log per session (idempotent via request_id)
• BoundaryResponse -- Structured fallback when semantic confidence is low (threshold: 0.45)
• ProvenanceChain -- Derivation trail attached to every tool response

Data Sources

Source	Entries	License
APQC PCF 7.4	1,921 processes	Royalty-free with attribution
ITIL 4	141 nodes	Practice names only (industry terms)
SCOR 12.0	164 nodes	Open-access standard
AI-era extensions	99 nodes	Original (MIT)
Total	2,325 processes
KPI metrics	3,910	From APQC PCF Metrics

Bilingual: Chinese (zh) + English (en).

Third-Party Attribution

APQC Process Classification Framework® (PCF) is an open standard developed by APQC, a nonprofit that promotes benchmarking and best practices worldwide. Used under APQC's royalty-free license for derivative works. To download the original PCF, visit apqc.org/pcf.

ITIL® is a registered trademark of PeopleCert group. This project references ITIL 4 practice names as industry-standard terminology. All descriptions are independently written and are not reproduced from ITIL publications.

SCOR® (Supply Chain Operations Reference) is a product of ASCM. This project references SCOR 12.0 process names as open-access industry-standard terminology. All descriptions are independently written.

Blueprint v1.0

The framework schema reserves five structural pillars for future enhancement:

Pillar	Status	Description
contract	Reserved	Process contract definitions
genome	Reserved	Process genome library
temporal	Reserved	Temporal patterns
interference	Reserved	Cross-process interference graph
outcomes	Reserved	Outcome dependency graph

These fields are present in schema.json as required empty defaults. Content will be populated in v2.0.

Development

# Install dependencies
uv sync

# Lint
ruff check .

# Test
pytest

# Full check (lint + test + benchmark)
ruff check . && pytest && pytest --benchmark-only

Project Structure

src/oprocess/
├── server.py              # FastMCP entry point (stdio/SSE/HTTP)
├── gateway.py             # ToolGatewayInterface + PassthroughGateway
├── auth.py                # Bearer token auth middleware
├── config.py              # pyproject.toml config loader
├── validators.py          # Input validation + sanitization
├── prompts.py             # 3 MCP prompt templates
├── tools/
│   ├── registry.py        # 7 tool registrations
│   ├── search.py          # search_process + map_role_to_processes
│   ├── resources.py       # 6 MCP resources
│   ├── export.py          # Responsibility document builder
│   ├── helpers.py         # Provenance + comparison utilities
│   ├── serialization.py   # ToolResponse → JSON
│   └── rate_limit.py      # Per-client rate limiter
├── governance/
│   ├── audit.py           # SessionAuditLog
│   ├── boundary.py        # BoundaryResponse
│   └── provenance.py      # ProvenanceChain
└── db/
    ├── connection.py       # SQLite + sqlite-vec connection
    └── queries.py          # All SQL queries

Tech Stack

• Runtime: Python 3.10+
• MCP Framework: FastMCP 2.x
• Validation: Pydantic 2.x (Annotated[..., Field(...)])
• Database: SQLite + sqlite-vec (optional vector search)
• Embeddings: gemini-embedding-001 (768-dim, optional — falls back to LIKE search)
• Packaging: uv + hatchling

License

MIT — applies to all source code and AI-era original content.

Third-party framework data (APQC PCF, ITIL, SCOR) is used under their respective licenses. See Third-Party Attribution for details.