Developers & AI Integration Hub
The authoritative technical portal for software engineers, AI architects, and DevOps leads. Integrate Tasleemat across Python applications, CLI automation pipelines, Model Context Protocol (MCP) servers, and autonomous multi-agent orchestration swarms.
📚 Dedicated Technical Guides & Modules
CLI Tooling & Scaffolding
Deep dive into tasleemat init, list, search, config, generate, and test. Interactive wizards, flags, air-gapped setups, and bash automation recipes.
Python SDK & Core Architecture
All 8 verified pure-Python functions, FastAPI integration, custom Streamlit dashboards, data models (ArtifactSummary, ArtifactAssets), and EVM calculation engine.
Model Context Protocol (MCP)
JSON-RPC 2.0 stdio server connecting Claude Desktop, Cursor, and AI agents directly to the 114 deliverables, schemas, and tailoring engine. Includes Python MCP client tutorials.
Draft 2020-12 & Self-Correction
114 companion JSON schemas, RFC 3339/4122/3986 format validation, and the autonomous iterative feedback healing loop for frontier LLMs with runnable retry scripts.
Agentic Swarms & CI/CD
Multi-agent PMO personas (CrewAI, LangGraph), tasleemat.ai.AIClient synthesis, headless GitHub Actions quality gates, and EVA math contracts.
1. Quick Installation & Environment Prerequisites
PyPI: tasleematTasleemat is distributed as a zero-dependency, pure-Python package on PyPI, supporting Python 3.9, 3.10, 3.11, 3.12, and 3.13. The core distribution packages all 114 bilingual deliverable templates, JSON companion schemas, CSV data dictionaries, CLI commands, and the Model Context Protocol (MCP) server.
# Install from PyPI
pip install tasleemat
# Verify both CLI and MCP executables
tasleemat --version
tasleemat-mcp --help🏛️ Tasleemat Architectural Blueprint
How developer tools, MCP servers, and autonomous agents interact with the canonical forms catalog and validation contracts:
flowchart TD
subgraph Clients["Engineering & AI Clients"]
CLI["Terminal CLI (tasleemat)"]
SDK["Python Application (tasleemat SDK)"]
MCP["AI IDE (Claude Desktop, Cursor, Antigravity)"]
Swarm["Autonomous Agent Swarm (LangGraph, CrewAI)"]
end
subgraph Core["Core Engine (tasleemat v2.5.0.post1)"]
Catalog["Catalog & Asset Resolver"]
Tailoring["Tailoring Engine (Tiers 1-3, 5 Packs)"]
Validator["Draft 2020-12 Schema Validator"]
Attester["Attested Math Engine (EVA PMO-06.05)"]
end
subgraph Data["Canonical Immutability Core"]
Forms["114 Deliverables (EN & AR)"]
Schemas["114 Draft 2020-12 Schemas"]
Dicts["114 CSV Dictionaries"]
end
CLI --> Catalog
SDK --> Catalog
MCP --> Catalog
Swarm --> Validator
Catalog --> Tailoring
Tailoring --> Forms
Validator --> Schemas
Attester --> Forms
2. CLI Command Reference
Dedicated CLI Page →
Scaffold customized workspaces (tasleemat init), search across English and Arabic deliverables (tasleemat search),
configure LLM credentials securely (tasleemat config), synthesize draft documents (tasleemat generate),
and run internal acceptance test suites (tasleemat test).
tasleemat init --tier 2 --pack agile --lang ar --code "PRJ-CLOUD" --name "مشروع السحابة" --out ./prj3. Python SDK Core Architecture
Dedicated SDK Page →
Pure-Python API exporting 8 verified functions: list_artifacts, get_artifact_definition,
get_artifact_assets, get_artifact_asset, get_tailoring_profile,
validate_artifact_data, get_lexicon, and render_artifact.
import tasleemat
profile = tasleemat.get_tailoring_profile(tier=2, pack="agile")
report = tasleemat.validate_artifact_data("PMO-03.01", {"project_id": "PRJ-01"}, mode="draft")4. Model Context Protocol (MCP) Server
Dedicated MCP Page →
Standard JSON-RPC 2.0 stdio server providing 6 specialized tools for Claude Desktop, Cursor, and AI agents:
tasleemat_list_forms, tasleemat_get_form, tasleemat_get_schema,
tasleemat_validate, tasleemat_tailor_project, and tasleemat_calculate_eva.
# Run standalone MCP server
tasleemat-mcp5. JSON Schemas & RFC Formats
Dedicated Schemas Page →All 114 deliverables feature companion JSON schemas conforming strictly to JSON Schema Draft 2020-12. Enforces strict format validations for RFC 3339 (dates/timestamps), RFC 4122 (UUIDs), RFC 3986 (URIs), and RFC 5322 (emails).
6. The Autonomous Agent Self-Correction Loop
Read Self-Correction Guide →
Autonomous agents don't guess: when LLMs generate invalid JSON or markdown tables, tasleemat.validate_artifact_data
returns exact JSON path errors. The agent incorporates the error message into its subsequent prompt to heal the output automatically.
7. Multi-Agent PMO Swarm Architecture
Read Multi-Agent Guide →Orchestrate specialized AI personas (Project Manager, PMO Governance Auditor, Risk & Ethics Lead, Executive Sponsor) using CrewAI or LangGraph while maintaining strict human-in-the-loop review boundaries before formal stage gates.
8. Programmatic AI Synthesis with tasleemat.ai.AIClient
Read AIClient Reference →
Use the unified tasleemat.ai.AIClient interface to synthesize deliverables across Google Gemini, OpenAI GPT-4o,
Anthropic Claude, and DeepSeek with standardized prompt assembly.
9. Headless CI/CD Automation & Quality Gates
Read CI/CD Guide →Integrate Tasleemat into GitHub Actions or GitLab CI to automatically enforce deliverable completeness, schema validity, and bilingual consistency on every commit or pull request.
10. Attested Computations & Deterministic Math Contracts (PMO-06.05)
Read Math Contracts →
AI models hallucinate mathematical figures. Tasleemat enforces deterministic calculation engines for Earned Value Analysis
(PMO-06.05) and Risk Severity matrices so project metrics remain mathematically attested.