TracePulse PLM Skill Packages product-line — 5-agent PLM PDF→BPMN chain, CSV→BPMN process_generator, output_format_controller, build_registry, plm_orchestrator (Wave 6.5 Conv A / PRD §3 #3). Full description in README.
Project description
plm-skill-packages
PRD §3 #3 product-line carve — Wave 6.5 Conv A.
Hosts the 5-agent PLM PDF→BPMN chain + CSV→BPMN generator + BPMN format controller + build registry + 5-agent orchestrator, plus skill manifests (YAML + filesystem) and the 4-state lifecycle FSM aligned verbatim with PRD §8.11.3.
Layout
plm_skill_packages/
├── plm_structural_analyst.py # PLM PDF→BPMN agent #1
├── plm_process_scout.py # PLM PDF→BPMN agent #2
├── plm_logic_modeler.py # PLM PDF→BPMN agent #3
├── plm_bpmn_transpiler.py # PLM PDF→BPMN agent #4
├── plm_quality_gate.py # PLM PDF→BPMN agent #5
├── process_generator.py # CSV→BPMN; only legitimate cascade fallback
├── output_format_controller.py # BPMN format correction + hidden retries
├── build_registry.py # File-handler dispatch
├── orchestrators/
│ └── plm_orchestrator.py # 5-agent chain coordinator
├── lifecycle.py # 4-state FSM (PRD §8.11.3 verbatim)
├── manifests.py # YAML manifest loader
├── registry.py # Skill manifest registry
├── contracts.py # Wire-contract Pydantic v2 frozen models
└── manifests/ # YAML skill manifests (one per skill)
├── plm.structural_analyst.yaml
├── plm.process_scout.yaml
├── plm.logic_modeler.yaml
├── plm.bpmn_transpiler.yaml
├── plm.quality_gate.yaml
├── plm.orchestrate_extraction.yaml
├── bpmn.generate_from_csv.yaml
└── bpmn.fix_format.yaml
Lifecycle FSM (Decision #167 — PRD §8.11.3 verbatim)
draft ──► released ──► deprecated ──► retired
│ │
└──────────────┴──► (terminal)
States are strictly forward-flowing (matching PRD §8.11.3 4-state capability lifecycle). No revivals, no skips. The FSM rejects any non-canonical transition.
Manifest store (Decision #168 — YAML + filesystem)
Manifests live under plm_skill_packages/manifests/*.yaml and load at boot via manifests.load_all(), matching the litellm_config/model_list.yaml catalog-as-truth pattern. No SQL, no admin UI in V1.
Re-export shims
Every monolith path that today references the lifted modules is preserved as a re-export shim — from agents.plm_structural_analyst import ... continues to work. Pattern matches Wave 1 W1.4 services/run_task_tracker.py (canonical decommission template).
Transitional dependency note
The lifted PLM agent + process_generator code keeps its existing from services.X / from models.X / from utils.X / from agents.base_agent imports. These resolve at runtime in the backend process via PYTHONPATH; standalone sibling tests configure conftest.py to extend sys.path. Wave 7 Extraction physically separates the repos and refactors inbound deps to ports/Protocols.
Wave 6.5 lineage
- Wave 6.5 Conv A (this) —
plm-skill-packagescarve. - Wave 6.5 Conv B —
plm-knowledgecarve (PRD §3 #6). - Wave 6.5 Conv C — closure audit + Wave 7 readiness re-author.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file plm_skill_packages-1.0.0.tar.gz.
File metadata
- Download URL: plm_skill_packages-1.0.0.tar.gz
- Upload date:
- Size: 71.5 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.14.0
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
6277c022cc9418350f316a86754e2d967b047b6285a14ff83a9ad4f070874250
|
|
| MD5 |
06bb79c66aa1af720a26b564d22374e0
|
|
| BLAKE2b-256 |
ddd339d18b998144878b3b2a94e5c0045930e6197360ebcb9cd37555c49c0d39
|
File details
Details for the file plm_skill_packages-1.0.0-py3-none-any.whl.
File metadata
- Download URL: plm_skill_packages-1.0.0-py3-none-any.whl
- Upload date:
- Size: 79.0 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.14.0
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
a0667e148717260408aa3e45f77e546089a3d03f44fd50b447cd0b1707755eb4
|
|
| MD5 |
aebb3b459998acb0366df00b955dec94
|
|
| BLAKE2b-256 |
874d74f3d2979014898ccb6b2b2cb6e1aa3c57c73d139302a56d445e3cd8a256
|